15 - Democracy in Hard Places-Oxford University Press (2022)

Democracy in Hard Places
Democracy
in Hard Places
Edited by
S C O T T M A I N WA R I N G A N D TA R E K M A S OU D
Oxford University Press is a department of the University of Oxford. It furthers
the University’s objective of excellence in research, scholarship, and education
by publishing worldwide. Oxford is a registered trade mark of Oxford University
Press in the UK and certain other countries.
Published in the United States of America by Oxford University Press

198 Madison Avenue, New York, NY 10016, United States of America.
© Oxford University Press 2022
All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted, in any form or by any means, without the
prior permission in writing of Oxford University Press, or as expressly permitted
by law, by license, or under terms agreed with the appropriate reproduction
rights organization. Inquiries concerning reproduction outside the scope of the
above should be sent to the Rights Department, Oxford University Press, at the
address above.
You must not circulate this work in any other form

and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data

Names: Mainwaring, Scott, 1954– editor. | Masoud, Tarek E., editor.
Title: Democracy in hard places / [edited by] Scott Mainwaring, Tarek Masoud.
Description: New York, NY : Oxford University Press, 2022. |
Includes bibliographical references and index.
Identifiers: LCCN 2022006297 (print) | LCCN 2022006298 (ebook) |
ISBN 9780197598764 (paperback) | ISBN 9780197598757 (hardback) |
ISBN 9780197598788 (epub)
Subjects: LCSH: Democracy—Case studies. | Democratization—Case studies. |
World politics—21st century—Case studies.
Classification: LCC JC423 .D381319 2022 (print) | LCC JC423 (ebook) |
DDC 321.8—dc23/eng/20220322
LC record available at https://lccn.loc.gov/2022006297
LC ebook record available at https://lccn.loc.gov/2022006298
DOI: 10.1093/oso/9780197598757.001.0001
1 3 5 7 9 8 6 4 2
Paperback printed by Lakeside Book Company, United States of America
Hardback printed by Bridgeport National Bindery, Inc., United States of America
Contents
List of Figures vii

List of Tables ix
Acknowledgments xi
Contributors xiii
Abbreviations xv
1. Introduction: Democracy in Hard Places 1

Tarek Masoud and Scott Mainwaring
2. India’s Democratic Longevity and Its Troubled Trajectory 34
Ashutosh Varshney
3. The Politics of Permanent Pitfalls: Historical Inheritances
and Indonesia’s Democratic Survival 73
Dan Slater
4. Africa’s Democratic Outliers: Success amid Challenges in Benin
and South Africa 94
Rachel Beatty Riedl
5. Georgia, Moldova, and Ukraine: Democratic Moments in
the Former Soviet Union 128
Lucan Ahmad Way
6. The Puzzle of Timor-Leste 160
Nancy Bermeo
7. Economic Crises, Military Rebellions, and Democratic
Survival: Argentina, 1983–2021 188
Scott Mainwaring and Emilia Simison
8. Why Democracies Survive in Hard Places 228
Scott Mainwaring
References 269
Index 303
List of Figures
1.1. Democracy in the world, 1900–present 2

1.2. Liberal democracy scores of the nine cases in this volume, 1974–2017 4
1.3. Marginal effects of continuous predictors 15
1.4. Marginal effects of qualitative predictors 17
1.5. India (1978–2017), probability of democratic breakdown (benchmarked
against Spain) 20
1.6. Indonesia (2000–2017), probability of democratic breakdown
(benchmarked against Spain) 22
1.7. Benin (1992–2017), probability of democratic breakdown (benchmarked
against Spain) 23
1.8. South Africa (1996–2017), probability of democratic breakdown
1.9. Georgia (2005–2017), probability of democratic breakdown
1.10. Ukraine (1995–1997, 2007–2013), probability of democratic breakdown
1.11. Moldova (1995–2004, 2010–2017), probability of democratic breakdown
1.12. Timor-Leste (2002–2017), probability of democratic breakdown
1.13. Argentina (1984–2017), probability of democratic breakdown
2.1. Electoral and liberal democracy index, India 1950–2020 38
2.2. The electoral-liberal democracy gap—decadal comparison 39
2.3. Electoral democracy index: India and some world regions, 1950–2020 42
2.4. Liberal democracy index: India and some world regions, 1950–2020 43
2.5. Electoral democracy index: India compared with selected countries,
1950–2020 44
2.6. Liberal democracy index: India compared with selected countries,
1950–2020 45
3.1. Indonesian economic growth rates, 1961–2019 77
viii list of Figures
3.2. Indonesian economic development levels, 1960–2019 78

3.3. Indonesia’s liberal democracy score over time, 1953–2019 79
4.1. Regime transition pathways 97
4.2. Pre-transition economic decline in South Africa and Benin, 1980–1994 102
4.3. V-Dem liberal democracy index rankings in South Africa and Benin,
1960–2019 105
6.1. Transition year GDP vs. mean for existing democracies, 1945–2015 161
6.2. Polity scores for former Portuguese colonies, 1975–2015 162
6.3. Polity scores for post-1989 Asian democracies, 1990–2017 162
6.4. Civil liberties for Southeast Asia and India, 1990–2017 163
6.5. Liberal democracy for Southeast Asia and India, 1990–2017 164
6.6. Conflict history and democratic durability, Kaplan–Meier Plot 165
6.7. Property rights in Timor-Leste vs. Indonesia, 1975–2015 170
7.1. Graphic summary of the argument 189
7.2. Argentina, V-Dem liberal democracy and Freedom House scores,
1983–2020 202
7.3. V-Dem illiberalism scores—FPV-PJ, UCR, and PRO, 1983–2019 209
8.1. Governing party illiberalism scores 234
8.2. Percentage of national assembly seats controlled by the governing party
and coalition 246
List of Tables
1.1. Regression Analysis of Correlates of Democratic Breakdown 12

1.2. Values for Each Case on Key Independent Variables (Socioeconomic
Indicators Averaged over Democratic Lifespan) 19
6.1. Aid, Peacekeeping, and Democratic Durability 181
8.1. Illiberalism Scores of Governing Parties and Regime Outcomes 239
8.2. Weighted Legislative Powers Scores for Nine Countries 255
8.3. Expert Perception of Judicial Capacity to Constrain Governments,
V-Dem (2019) 256
8.4. World Bank Governance Indicators: Perceptions of Control of Corruption
and Rule of Law in Nine Hard Cases, 2019 257
8.5. Perception of Government Effectiveness and Per Capita GDP Growth in
Nine Hard Cases 259
Acknowledgments
This book is about why democracy sometimes survives for a long time in difficult
conditions. When we began planning the conference that led to the book, the
COVID-19 pandemic and the January 6 insurrection at the United States Capitol
would have been unthinkable to all but the most determined pessimists, but it
was already a time of grave worry about the state of democracy in the world. We
hope that this book not only enriches our understanding of what makes democ-
racy in hard places possible, but also that it might inspire belief during these hard
times that democracy can persist in the face of grave challenges.
The idea for this volume grew out of many conversations we had from 2016
to 2019 at Harvard University. We had neighboring offices, co-taught a course
on “Getting and Keeping Democracy” in 2017 and in 2018 at Harvard’s John
F. Kennedy School of Government, and co-directed a program on “Democracy
in Hard Places” at the Kennedy School’s Ash Center for Democratic Governance
and Innovation. During these years of working together, we often disagreed
with one another—but in so doing, we always pushed each other to new
understandings and insights. We also accrued a number of debts that we are
pleased to acknowledge here.
Our first debt is to Tony Saich, the director of the Ash Center. Tony and the
Center generously supported our work from the beginning. This support made
possible the eponymous conference in May 2019 that gave rise to this book, and
also enabled us to secure the editorial and administrative assistance necessary to
bring the book to fruition. Special thanks must be rendered to Melissa D’Anello
and Maureen Griffin of the Ash Center, without whom neither the conference
nor the book would have happened. We also thank our assistants at the Kennedy
School, Juanne Zhao and Sari Betancourt.
Our debts were not just financial and administrative, but intellectual. We, and
the contributors to this volume, were fortunate to benefit from the insights of an
outstanding group of commentators and participants in the May 2019 conference
on Democracy in Hard Places, including Eva Bellin, Fernando Bizzarro, Melani
Cammett, Timothy Colton, Steven Fish, Candelaria Garay, Frances Hagopian,
Sarah Hummel, Steven Levitsky, Pia Raffler, and Deborah Yashar. We are also
grateful to Harvard colleagues Daniel Ziblatt, Richard Zeckhauser, Gautam Nair,
Moshik Temkin, Alex Keyssar, and Archon Fung for informal conversations that
shaped our thinking about democracy and its survival. Ashutosh Varshney, the
author of the chapter on India, deserves special thanks for being an invigorating
xii Acknowledgments
intellectual presence at the Ash Center during a critical phase of this book’s
development.
We thank David McBride of Oxford University Press for his enthusiastic sup-
port for the project from the outset. And we record our thanks to María Victoria
De Negri who, with characteristic talent and attention to detail, prepared the
book for publication and compiled the index. Fernando Bizzarro provided
helpful research.
Finally, Scott thanks his wonderful partner, Sue Elfin, for gracing his life.
Tarek thanks his long-suffering partner, Kristin Alcorn Masoud, for tolerating
his frequent failures to bring grace to hers.
The authors dedicate this book to the memory of Alfred C. Stepan, whose
fingerprints are on every chapter.
Contributors
Nancy Bermeo (PhD Yale University) is currently a Nuffield Senior Research Fellow at
Oxford University and a PIIRS Senior Scholar at Princeton University. She writes on the
causes and consequences of political mobilization and regime change as well as the quality
of democracy. Her books include an award-winning study of the breakdown of democ-
racy titled Ordinary People in Extraordinary Times (Princeton University Press), and, with
Deborah Yashar, Parties, Movements and Democracy in the Developing World (Cambridge
University Press), plus Mass Politics in Tough Times with Larry Bartels (Oxford University
Press), and Coping with Crisis: Government Reactions to the Great Recession with Jonas
Pontusson (Russell Sage Foundation). Her latest book, Democracy after War, is forth-
coming from Princeton University Press.
Scott Mainwaring (PhD, Stanford) is the Conley Professor of Political Science at the
University of Notre Dame. His research and teaching focus on democratization, party
systems, and Latin American politics. His book with Aníbal Pérez-Liñán, Democracies
and Dictatorships in Latin America: Emergence, Survival, and Fall (Cambridge University
Press, 2013) won the Best Book Prizes of the Democracy and Autocracy section of the
American Political Science Association and the Political Institutions section of the Latin
American Studies Association. Mainwaring was elected to the American Academy of
Arts and Sciences in 2010. In April 2019, PS: Political Science and Politics listed him as
one of the fifty most cited political scientists in the world. He served as the Jorge Paulo
Lemann Professor for Brazil Studies and as faculty co-chair of the Brazil Studies program
at Harvard University from 2016 to 2019.
Tarek Masoud (PhD Yale University) is the Ford Foundation Professor of Democracy and
Governance at Harvard University’s John F. Kennedy School of Government. He is the
co-Editor of the Journal of Democracy, the director of the Kennedy School’s Middle East
Initiative, the Initiative on Democracy in Hard Places, and the co-author of, among other
works, The Arab Spring: Pathways of Repression and Reform (Oxford University Press, 2015).
Rachel Beatty Riedl (PhD Princeton University) is the John S. Knight Professor of
International Studies, Director of the Einaudi Center for International Studies, and a pro-
fessor in the Department of Government at Cornell University. Her research interests
include institutional development in new democracies, local governance and decentral-
ization policy, authoritarian regime legacies, and religion and politics, with a regional
focus in Africa. She is the author of award-winning Authoritarian Origins of Democratic
Party Systems in Africa (Cambridge University Press, 2014) and co-author with Gwyneth
McClendon of From Pews to Politics: Religious Sermons and Political Participation in
Africa (Cambridge University Press, 2019). Riedl is co-host of the podcast Ufahamu
Africa, featuring weekly episodes of news highlights and interviews about life and politics
on the African continent.
xiv Contributors
Emilia Simison (MA Torcuato Di Tella University) is a PhD candidate in Political

Science at the Massachusetts Institute of Technology. Her research focuses on the com-
parative political economy of policymaking and policy change, especially on how po-
litical institutions in democratic and authoritarian regimes shape the extent to which
citizens and interest groups influence policy. Prior to MIT, she was a PhD fellow at the
Argentine National Scientific and Technical Research Council (CONICET) working at
Gino Germani Research Institute, and she taught at the University of Buenos Aires and
Torcuato Di Tella University.
Dan Slater (PhD Emory University) is Weiser Professor of Emerging Democracies and
Director of the Weiser Center for Emerging Democracies (WCED) at the University
of Michigan. He specializes in the politics and history of enduring dictatorships and
emerging democracies, with a regional focus on Southeast Asia. He previously served
as Director of the Center for International Social Science Research (CISSR), Associate
Professor in the Department of Political Science, and associate member in the Department
of Sociology of the University of Chicago. He is the author of Ordering Power: Contentious
Politics and Authoritarian Leviathans in Southeast Asia (Cambridge University Press,
2010) and co-author of Coercive Distribution (Cambridge University Press, 2018).
Ashutosh Varshney (PhD Massachusetts Institute of Technology) is Sol Goldman

Professor of International Studies and the Social Sciences and Professor of Political
Science at Brown University, where he also directs the Center for Contemporary South
Asia. Previously, he taught at Harvard, Notre Dame, and the University of Michigan,
Ann Arbor. His books include Battles Half Won: India’s Improbable Democracy (Penguin
Books); Collective Violence in Indonesia (Lynne Rienner Publishers); Ethnic Conflict
and Civic Life: Hindus and Muslims in India (Yale University Press); India in the Era of
Economic Reforms (with Jeffrey Sachs and Nirupam Bajpai, Oxford University Press);
and Democracy, Development, and the Countryside: Urban–Rural Struggles in India
(Cambridge University Press).
Lucan Ahmad Way (PhD University of California, Berkeley) is Professor of Political

Science at the University of Toronto. Way’s research focuses on democratization and
authoritarianism in the former Soviet Union and the developing world. His most re-
cent book (with Steven Levitsky), Social Revolution and Authoritarian Durability in the
Modern World (forthcoming from Princeton University Press), provides a comparative
historical explanation for the extraordinary durability of autocracies born of violent so-
cial revolution. Professor Way’s solo authored book, Pluralism by Default: Weak Autocrats
and the Rise of Competitive Politics (Johns Hopkins University Press, 2015), examines
the sources of political competition in the former Soviet Union. His book Competitive
Authoritarianism: Hybrid Regimes after the Cold War (with Steven Levitsky), was
published in 2010 by Cambridge University Press. Way’s work on competitive authori-
tarianism has been cited thousands of times and helped stimulate new and wide-ranging
research into the dynamics of hybrid democratic-authoritarian rule.
Abbreviations
ADP Agrarian Democratic Party

AEI Alliance for European Integration
AITI Association for the Integration of Timor in Indonesia
ANC African National Congress
APODETI Popular Democratic Association of Timor
ASDT Timorese Social Democratic Association
BDPM Bloc for a Democratic and Prosperous Moldova
BJP Indian People’s Party
BPP European Solidarity /Petro Poroshenko Bloc
BYuT Yulia Tymoshenko Bloc
CA Constituent Assembly (India)
CAA Citizenship Amendment Act
CAC Argentine Chamber of Commerce
CDM Electoral Bloc Democratic Convention of Moldova
CEMIDA Center of Military Members for the Argentine Democracy
CGT General Confederation of Labor
CNRT National Congress for Timorese Reconstruction
CONADEP National Commission on the Disappearance of Persons
COSATU Congress of South African Trade Unions
CRIET Court of Punishment of Economic Crimes and Terrorism
DPM Democratic Party of Moldova
ECLAC United Nations Economic Commission for Latin America and the
Caribbean
ENM United National Movement
ESMA School of Mechanics of the Navy
EU European Union
F-FDTL Timor-Leste Defense Force
FALINTIL Armed Forces for the National Liberation of Timor-Leste
FARD Action Front for Renewal and Development
FCBE Cowry Forces for an Emerging Benin
FPV-PJ Front for Victory-Peronist Party
FRELIMO Mozambique Liberation Front
FRETILIN Revolutionary Front for an Independent Timor-Leste
FRTLI Revolutionary Front for an Independent East Timor
GD Georgian Dream
GDP Gross Domestic Product
Gerindra Great Indonesia Movement Party
GNI Gross National Income
xvi Abbreviations
GNU Government of National Unity

GOLKAR Party of Functional Groups
IMF International Monetary Fund
INC Indian National Congress
JD Janata Dal
JNP Janata Party
KGB Committee for State Security
KOTA Association of Timorese Heroes
MODIN Movement for Dignity and Independence
MPLA Popular Movement for the Liberation of Angola
NDPU People’s Democratic Party of Ukraine
NF People’s Front
NGO nongovernmental organization
NP National Party
NRC National Registry of Citizens
NU Nahdlatul Ulama
NUNS Our Ukraine‚ People’s Self-Defense Bloc
OAS Organization of American States
OIC Organization of Islamic Cooperation
OECD Organization for Economic Co-operation and Development
OPEC Organization of the Petroleum Exporting Countries
PAN National Mandate Party
PAS Party of Action and Solidarity
PCRM Party of Communists of the Republic of Moldova
PD Democratic Party (Indonesia)
PD Democratic Party (Timor-Leste)
PDAM Agrarian Democratic Party of Moldova
PDI Indonesian Democratic Party
PDIP Indonesian Democratic Party of Struggle
PDM Democratic Party of Moldova
PIL Public Interest Litigation
PJ Peronist Party
PKB National Awakening Party
PKI Indonesian Communist Party
PLDM Liberal Democratic Party of Moldova
PLP People’s Liberation Party
PNI Indonesian Nationalist Party
PNTL National Police of Timor-Leste
PPP United Development Party (Indonesia)
PPP purchasing power parity
PR Party of Regions
PRD Democratic Renewal Party
PRO Republican Proposal
PRPB Peoples’ Revolutionary Party of Benin
Abbreviations xvii
PSD Social Democrat Party

PU Progressive Union
RB Renaissance Party of Benin
RENETIL National Resistance of East Timorese Students
RUKH People’s Movement of Ukraine
SACP South African Communist Party
SICONARA Ship Technicians’ Union
SJP(R) Samajwadi Janata Party (Rashtriya)
SN Servant of the People
SRA Argentine Rural Society
TMC Trinamool Congress
TRC Truth and Reconciliation Commission
UAE United Arab Emirates
UCEDE Union of the Democratic Center
UCR Radical Civic Union
UDT Timorese Democratic Union
UIA Argentine Industrial Union
UN United Nations
UNAMET United Nations Mission to East Timor
UNDERTIM National Union of the Timorese Resistance
UNM United National Movement
UNSD United Nations Statistics Division
UNTAET United Nations Transitional Administration in East Timor
UP Uttar Pradesh
USAID United States Agency for International Development
ZYU For a United Ukraine!
1
Introduction
Democracy in Hard Places
Tarek Masoud and Scott Mainwaring
If recent political events have taught us anything, it is that democracy is often

fragile. Throughout the world, in such places as Hungary, Poland, India, and
Brazil, democratic regimes now find themselves imperiled by the rise of ultra-
nationalist and populist leaders who pay a steady lip service to the will of the
people while daily undermining freedom, pluralism, and the rule of law. Not
even the wealthiest and most powerful of the world’s democracies—the United
States of America—has proven immune. The one-time “arsenal of democracy”
is now sometimes held up as a candidate for democratic backsliding. In their
2018 bestseller, How Democracies Die, Steven Levitsky and Daniel Ziblatt tes-
tify to an “epidemic of norm breaking that now challenges our democracy” and
warn of an American future in which no-holds-barred partisan warfare leads to
either a perpetual state of crisis or the inauguration of a full-blown, one-party
regime. As if in agreement, Freedom House now ranks the United States sixty-
first out of 210 countries in terms of its level of freedom, behind much younger
democracies such as the Czech Republic, Slovakia, Lithuania, Greece, Chile, and
Taiwan (Freedom House, n.d.). Although we do not agree with this judgment,
the Polity V project assigns the United States a score of 8 out of 10 for the year
2016, lower than its score from 1809–50 (Marshall and Jaggers 2020). And in
2015, the Economist Intelligence Unit, which maintains its own “Democracy
Index,” downgraded the United States to a “flawed democracy,” a category re-
served for countries with free and fair elections and basic civil liberties but with
“problems in governance, an underdeveloped political culture and low levels of
political participation” (The Economist Intelligence Unit 2020, 53).
An illustration of how much democracy has lapsed, even in places where
we would not have expected it to, can be seen in Figure 1.1, which plots one
hundred years of global and OECD average scores on the “liberal democracy”
index compiled by the University of Gothenburg’s Varieties of Democracy
project (Coppedge et al. 2021). Since peaking around 2011, average scores
on that index—which captures the extent of civil liberties, rule of law, judicial
2 Democracy in Hard Places
.8
Liberal Democracy Index (V−Dem, v11)

.6
.4
.2
0
1900 1920 1940 1960 1980 2000 2020
Year
OECD (population weighted)

World (population weighted)
Figure 1.1. Democracy in the world, 1900–present

Source: Coppedge et al. 2021.
independence, checks on executive power, and electoral integrity—have now

declined to levels not seen since the end of the Cold War, and the trend appears
to point downward. “Nor,” as Larry Diamond has written, “do the numbers
capture the full extent of the danger.” According to Diamond, “China, Russia,
and their admirers are making headway with a new global narrative, hailing
strongman rule—not government by the people—as the way forward in dif-
ficult times.”1 The current global pandemic threatens to make a bad situation
worse, with increasing unemployment and, in many countries, shrinking
GDPs fueling popular anger and testing the limits of mass and elite faith in
democracy in ways that appeared to accelerate democratic regressions around
the world.
These new doubts about democracy’s survival in some of the world’s most
prosperous countries lend new relevance to studies of its survival in places less
blessed by abundance. After all, if American democracy can be compromised
by heightened polarization and the willingness of some political elites to un-
dermine long-standing norms in efforts to gain electoral advantages, then
democracy’s persistence in large, poor, ethnically diverse countries such as
India and Indonesia, with their low rates of educational attainment, rickety
governing apparatuses, and frequent economic difficulties, is nothing short
of miraculous. Moreover, taking stock of democracy’s record across time and
space, we find several such instances of democratic survival in the face of
seemingly long odds. Since 1983, Argentines have held onto their democracy
Introduction 3
despite three protracted economic depressions, military mutinies, and a prior

history of repeated democratic failures. In the 1990s, South Africans built one
of Africa’s longest-lived democracies out of the ruins of a system of white su-
premacy that dehumanized the great majority of citizens and pitted the races
against each other. And in the early 2000s, the people of Timor-Leste erected a
multi-ethnic democracy after a decades-long civil war that laid waste to count-
less lives and livelihoods. These cases demand our attention, both because they
too face ever-present risks of democratic decay and downfall and because we
may be able to derive from them lessons about how democracy can be fortified
in times of challenge.
This, then, is a book about how democracy persists when all signs suggest
that it should not. It puts front and center cases of what we call “democracy in
hard places”: countries that lack the structural factors and exist outside of the
contexts that scholars have long associated with democracy’s emergence and
endurance. Democracies in hard places overcome underdevelopment, ethno-
linguistic diversity, state weakness, and patriarchal cultural norms. They tame
grasping, politically ambitious militaries; transcend influences and pressures
from autocratic neighbors; and cope with polarized political parties. Without
denying that democracy is easier to build and hold onto in societies that are
free of such hurdles, this book asks what we can learn about strengthening de-
mocracy from those that managed to leap over them. By theorizing about dem-
ocratic survival from such cases—which, in the parlance of social science, lie
“off the regression line”—we capture what Michael Coppedge identifies as “the
greatest potential to innovate and challenge old ways of thinking” (Coppedge
2002, 16). Are democracies in hard places the equivalent of lottery winners—
dramatic exceptions to fundamental rules? Or is there something systematic
that can be gleaned from such cases about how democracy can be erected and
upheld around the world?
To answer these questions, this book presents nine case studies—written by
leading experts in the discipline—of episodes in which democracy emerged
and survived against long odds. The cases are drawn from almost every re-
gion of the world that formed part of what Samuel Huntington called the
“third wave” of democracy, which began in southern Europe in the mid-1970s,
spread to Latin America in the late 1970s and 1980s, and to Eastern Europe
and sub-Saharan Africa in the 1990s. Six of the cases are ones of long-term
democratic survival—Argentina (1983–present), Benin (1991–2019), India
(1977–present), Indonesia (1999–present), South Africa (1995–present), and
Timor-Leste (2002–present). The other three have more mixed democratic
records—Georgia (2005–present), Moldova (1995–2005, 2010–present), and
Ukraine (1995–98, 2007–14). In each case, many of the conditions convention-
ally associated with durable democracy were either attenuated or absent. Each
case study details the constellation of obstacles to democracy faced by a given

country, describes the major political actors with the potential to impact re-
gime trajectories, and explains how the threat of democratic breakdown was
staved off or averted.
Figure 1.2 plots the V-Dem “liberal democracy” scores of the cases in this
book from 1974 to 2017 (the most recent year available).
Although the case studies presented in this book do not offer a unified answer
to the question of how democracy survives in inauspicious conditions, readers
will find in them powerful rejoinders to structural accounts of democracy’s
emergence and survival. This is perhaps to be expected, given that the cases
were selected based on their want of democracy’s hypothesized structural
causes. But, as the editors of this volume, we find in the narratives of demo-
cratic survival offered in this book striking evidence of the importance of po-
litical actors, their normative beliefs, and how these beliefs shape their choices.
Twenty-five years ago, Adam Przeworski and Fernando Limongi (1997, 176–77)
complained that, in structural accounts of democratization, “no one does an-
ything to bring democracy about; it is secreted by economic development and
the corollary social transformations.” In reality, they argued, “democracy was
an outcome of actions, not just of conditions.” The cases in this volume validate
that observation.
.7
.6
Liberal Democracy Index
.5
.4
.3
1980 1990 2000 2010 2020

Year
Argentina Benin Timor−Leste

Georgia India Indonesia
Moldova South Africa Ukraine
Figure 1.2. Liberal democracy scores of the nine cases in this volume, 1974–2017
Source: Coppedge et al. 2018.
Introduction 5
Agents, Interests, and Regime Preferences
Agent-centric accounts of political phenomena have sometimes taken the form

of “great person” narratives in which individual leaders operate with consid-
erable autonomy, and in which outcomes are the result of individual decision-
making. Without denying that individual leaders occasionally make a decisive
difference in outcomes, this book eschews that tendency. Its focus is not on
individuals but on key collective actors who command political resources and
influence political competition (Mainwaring and Pérez-Liñán 2013, 5). These
include chief executives and their ruling parties, other political parties, mili-
taries and coercive agencies, labor unions, social movements, and civil society
groups (Bermeo and Yashar 2016a). These are the agents whose preferences and
choices determine whether a democracy endures, or joins the long list of dem-
ocracies undone by military takeovers, auto-golpes, palace coups, and civil wars.
Second, each chapter takes seriously the constraints under which actors operate.
These constraints may take the form of structural or cultural conditions, formal
institutions (such as constitutional structures or electoral rules), international
actors, or other domestic actors (such as courts, militaries, or political parties).
In addition to these “external” constraints, many of the chapters also highlight
“internal” constraints that have hitherto received short shrift in accounts of dem-
ocratic emergence and survival—in particular, the existence of deeply rooted
normative commitments that shape the behavior of collective actors and take
certain courses of action off the table.
Thus, if one goal of this book is to refocus the study of democratic survival
from structural and external contextual conditions to the role played by polit-
ical actors, another is to contribute to the debate about what drives those actors.
What causes, for instance, a chief executive to forego an opportunity to expand
his power at the expense of democratic legitimacy? Why would a powerful mili-
tary in a fledgling democracy stick with a messy democratic system instead of an-
swering calls to enter the fray in the name of stability? Why would an incumbent
facing electoral defeat accept that defeat instead of tampering with the election
or simply abrogating it outright? Assuming that democratic survival can reason-
ably be thought of as a function of such choices, what explains those choices?
The literature has offered two broad answers to this question. The first holds
that actors’ preferences over regime type are driven purely by considerations of
self-interest. That is, they weigh the power and income they can expect to derive
from different governing arrangements and choose accordingly. Thus, Adam
Przeworski (1991) argues that democracy survives (i.e., is an equilibrium) when
the losers in a particular election believe that they have enough of a chance of
winning in the future to make sticking with democracy an attractive proposi-
tion.2 Such accounts assume, as Przeworski (2019, 172) does, that “[t]‌he dream
of all politicians is to remain forever in office and to use their tenure to do what-
ever they want.” In this telling, political actors accept democracy and the associ-
ated constraints on power only because it serves their interests, or because they
have no alternative.
The second answer to the question of why some political actors sustain de-
mocracy despite pressures to the contrary holds that self-interest or external
constraints are insufficient to explain most instances of democratic persist-
ence. In an earlier work, one of the authors of the present chapter points out
that parties and movements have often upheld democratic regimes despite
having plausibly better prospects outside of them: “Because actors believe in
the system, they are willing to make concessions to abide by the rules of the
game” (Mainwaring 1989). Without denying that actors also have instrumental
goals, this alternative view locates the sources of actors’ democracy-sustaining
decisions in their values and normative commitments. Canonical works that
share this view include Berman (1998), Bermeo (1992), Dahl (1971, 124–89),
Hofstadter (1969), Lepsius (1978), Levitsky and Ziblatt (2018), Linz (1978a,
1978b), and McFaul (2002). Garrard’s (2002) account of the emergence of de-
mocracy in Great Britain similarly emphasizes the acceptance of liberal values
by key collective actors. Along related lines, an extensive body of literature on
populism and the threat it poses to democracy hinges explicitly on populist
leaders’ hostility to the liberal aspects of democracy (Müller 2016b; Weyland
2020; Weyland and Madrid 2019). In short, these works conclude that democ-
racy survives when the principal political actors who are in a position to main-
tain a regime or tear it down bear a preference for democracy that extends
beyond self-interest and usually leads them to prioritize it over other impor-
tant desiderata.3
The chapters in this volume take different perspectives on this issue, and this
book is unlikely to resolve such a long-standing debate. At the same time, much
is to be gained by placing this question firmly on the agenda for both scholars
and practitioners. The urgency of understanding how political actors can be
made to sustain democracy in hard places cannot be exaggerated. It derives not
just from the fact that democracy around the world is under threat, but from
the realization that the greatest potential for democratic expansion lies in hard
places similar to the cases explored here. A recent survey by Mainwaring and
Bizzarro identified 103 transitions to democracy between 1975 and 2012, many
of which were in hard places, and almost half of which (46) broke down, with
an average democratic duration among the breakdown cases of only eight years
(Mainwaring and Bizarro 2020). If this statistic is to be improved, scholars and
practitioners must understand the factors that have enabled a small but not in-
significant number of hard places to transcend the constraints of structure and
achieve and maintain decent, representative government.
Introduction 7
The remainder of this chapter proceeds as follows. First, we offer a review of

some of the most important structural and contextual factors that have been
hypothesized to determine a country’s democratic prospects. We then identify
the nine cases whose democratic trajectories this book examines, explain why we
consider them hard places, and discuss how each of them fits into the book. We
end the chapter with the volume’s conclusions and contributions to the theory
and practice of democratization.
The Effects of Context
Democracy’s survival in “hard” places, and its looming retrenchment in com-

paratively “easy” ones, are two of the most powerful empirical rejoinders to the
structural and contextual accounts of democratic transition and democratic
stability. However, there is a reason those accounts became nearly hegemonic
in the field of comparative democratization—they have been validated by a
wealth of cross-national, statistical analyses. Accordingly, this project is predi-
cated on the well-grounded idea that democracy is more likely to survive under
some conditions than others. In this section, we discuss seven of these facili-
tating conditions: (1) economic development, (2) macroeconomic performance,
(3) economic inequality (or, more properly, the lack thereof), (4) state capacity,
(5) ethnic homogeneity, (6) democratic cultures, and (7) a pro-democratic in-
ternational environment. We then introduce the cases that demonstrate the
possibility for democratic survival even in the absence of some or all of these
hypothesized conditions.
Of all the hypothesized enabling conditions for democracy, perhaps the
one on which there is the broadest agreement is the achievement of a certain
kind of prosperity, built on a shift from agriculture to industry. More than sixty
years ago, Seymour Martin Lipset (1959, 75) famously found that “the average
wealth, degree of industrialization and urbanization, and level of education is
much higher for the more democratic countries” and concluded that “the more
well-to-do a nation, the greater the chances that it will sustain democracy.” In
the intervening decades, scholars confirmed Lipset’s initial empirical finding and
offered a variety of explanations for it (Boix 2011; Boix and Stokes 2003; Epstein
et al. 2006; Przeworski et al. 2000).4 Some have argued that affluent societies are
more capable of sustaining democracy because they have more educated citi-
zenries, who are thus more rational and tolerant and less receptive to the rad-
ical ideologies and emotional appeals of would-be autocrats. For others, the
correlation between economic development and democracy is largely due to the
growth, in prosperous societies, of autonomous civic groups that provide citi-
zens with the wherewithal to demand greater liberties and to deter or beat back
encroachments upon them. For still others, democracy follows from develop-
ment because the relative prosperity of the average citizen in a wealthy society
means that economic elites possess less reason to fear expropriation if they grant
or expand the franchise (Acemoglu and Robinson 2006; Boix 2003). Without
development, the reasoning goes, these and other beneficial mechanisms would
not operate, and political life would regress to a repressive and conflict-ridden
mean. As Alemán and Yang (2011, 1143) concluded in a study of democratic
transition and breakdown from 1970 to 1999, “the greatest threat to democratic
survival is a low level of income.”
Other scholars have noted that sustained periods of poor economic perfor-
mance, in which GDP per capita either declines or stagnates, can seriously lower
the probability of democratic survival (Gasiorowski 1995; Przeworski et al.
2000; Sing 2010; Svolik 2015). “Like war and revolution,” Wibbels and Roberts
(2010, 384) wrote, economic crises are “periods of severe disequilibria” that can
“transform a political landscape in profound ways,” and “shake the foundations
of nations.” The breakdown of Germany’s Weimar Republic in 1933 is the most
famous example in which widespread immiseration resulting from economic
crisis contributed to the end of a democratic regime. In their study of democracy
in twenty-six developing countries, Diamond, Lipset, and Linz (1987, 8) declared
“it is clear that, over the long term in particular, the effectiveness of democratic
regimes in satisfying people’s wants heavily affects their stability” and find that
the most successful democracies “have generally experienced relatively steady
economic growth, which in turn has strengthened their legitimacy.” It makes de-
ductive sense that citizens would care more about economic security and less
about democracy in moments of economic hardship. For this reason, Przeworski
(1991, 136) argued that there were steep tradeoffs between economic reform
and democracy in the post-communist countries and Latin America. Economic
reforms, he wrote, “are socially costly and politically risky. . . . [T]‌hey hurt large
social groups and evoke opposition from important political forces. And if that
happens, democracy may be undermined or reforms abandoned, or both.”
Yet another economic factor broadly associated with democracy’s emer-
gence and survival is the level of economic inequality. Carles Boix (2003) argues
that democracy emerges when inequality decreases: “As the distribution of in-
come becomes more equal among individuals, redistributive pressures from
the poorest social sectors on the well-off voters diminish. Accordingly, the rel-
ative costs [to the wealthy] of tolerating a mass democracy decline” (Boix 2003,
10). Related arguments have been made by Lipset (1959, 83) and Acemoglu and
Robinson (2006).5 In contrast, Ansell and Samuels (2010, 1545) predict a posi-
tive relationship between democratization and rising income inequality, as dem-
ocratic reforms are often demanded by “newly wealthy economic groups” that
“accumulate an increasing share of national income” that they seek to defend
against state predation. However, an empirical analysis by Haggard and Kaufman
Introduction 9
(2012) of third-wave democratic transitions and reversals finds little support for
the existence of a systematic relationship between inequality and democratic
emergence and survival.
A fourth, non-economic structural factor sometimes identified as important
to sustaining democracy is the strength of the state apparatus. At the extreme,
any regime—democratic or not—that cannot maintain what Max Weber called “a
monopoly over the legitimate use of violence” is unlikely to survive for long. But
even states that can manage this basic, minimal function of statehood may prove
unable to hang onto democracy if they lack the infrastructural power (Mann
1984) to carry out essential government functions of law enforcement, taxation,
and service delivery. O’Donnell (1993) describes the danger to new democracies
of “brown areas” in which state incapacity leads to a functional truncation of cit-
izenship rights. Slater, Smith, and Nair (2014, 354) argue that democracies with
ineffectual governing apparatuses quickly find themselves the objects of mass
discontent and highly susceptible to military intervention: “[A]‌dministrative in-
capacity means that recurrent crises of governability will repeatedly tempt and
enable military intervention to restore political stability. Meanwhile, democracy’s
chronic failure to ‘deliver the goods’ in weak-state settings will give the poor ma-
jority little reason to defend democracy against its enemies.”
A fifth contextual underpinning of democratic survival identified by scholars
is the presence of relative ethnic, linguistic, and religious homogeneity (Dahl
1971; Horowitz 2000; Rabushka and Shepsle 1972; Welsh 1993). Societies riven
by identity-based cleavages are thought to lack the comity that helps to keep dem-
ocratic competition from spilling over into bloodshed. As Arend Lijphart wrote
more than forty years ago, the difficulty of “achiev[ing] and maintain[ing] stable
democratic government in a plural society is a well-established proposition in
political science” (1977, 1). Although recent analyses of cross-national data have
attenuated the link between diversity and democratic survival (Fish and Brooks
2004; Teorell 2010), there nonetheless remain strong theoretical reasons why we
might expect ethnic diversity to inhibit democratic survival. An early expression
of these reasons was offered by John Stuart Mill, who, writing in the middle of
the nineteenth century, declared that “free institutions are next to impossible in a
country made up of different nationalities.” Divided by language and culture, he
reasoned, citizens in such societies eye each other with suspicion, as “one section
does not know what opinions, or what instigations, are circulating in another.”
Instead of uniting to hold national leaders accountable and constrain their
power, said Mill, the diverse citizenry’s “mutual antipathies are generally much
stronger than dislike of the government,” which, presumably, is able to divide
and conquer (Mill 1861, 289). A practitioner’s corroboration of the importance
of homogeneity to democracy was offered by the early twentieth-century British
prime minister, Arthur Balfour, who is supposed to have said, “[O]‌ur whole po-
litical machinery presupposes a people so fundamentally at one that they can
safely afford to bicker” (quoted in Friedrich 1939, 571). In the absence of that
salutary oneness, the reasoning goes, paper stones would soon be converted into
real ones. Scholars have found this danger to be particularly in a democracy’s
early stages, when political entrepreneurs might be most tempted to mobilize
ethnic sentiments in order to acquire power (Horowitz 2000; Snyder 2000).
A sixth theorized requisite of democracy is the existence in a given society of
what might be called a democratic “culture.” As Welzel and Inglehart (2009) put
it, “[M]‌ass beliefs are of critical importance for a country’s chances to become and
remain democratic.” Perhaps not surprisingly, for many scholars, the culture most
conducive to democracy is that of the Western countries in which modern forms
of mass, democratic government first emerged. Thus, Almond and Verba (1963,
5), argue that a “democratic form of participatory political system requires as well
a political culture consistent with it” and lament the difficulty of transferring “the
political culture of the Western democratic states to the emerging nations.” In his
study of democratization in Britain, Garrard (2002) describes a gradual process in
which previously disenfranchised groups acquired the right to participate upon
their achievement of a certain level of what he calls “political fitness,” defined largely
in terms of their attachment to classically liberal values held among the British elite.
Democracy’s early emergence and survival in Northern European coun-
tries caused some to implicate the distinctive religion of that part of the world,
Protestant Christianity. Woodberry (2012, 245) notes that “by World War I every
independent, predominantly Protestant country was a stable democracy—with
the possible exception of Germany,” a fact he attributes to Protestantism’s em-
phasis on mass education and religious liberty. Other forms of Christianity, in
contrast, were deemed more likely to uphold traditional, autocratic politics. For
instance, Lipset (1959, 93) declared that Roman Catholicism’s insistence of a mo-
nopoly on truth was at odds with democracy, “which requires, as part of its basic
value system, the belief that ‘good’ is served best through conflict among op-
posing beliefs.” Orthodox Christianity’s hospitality to democracy has similarly
been thrown into doubt (Marsh 2005), and Prodromou (2004, 62) has pointed
out that “Orthodox churches often display a certain ambivalence about key
elements of the pluralism that characterizes democratic regimes.” If the exist-
ence today of a large number of Catholic-majority democracies appears to falsify
Lipset’s hypothesis, democracy continues to be weak in the Eastern Orthodox
world (which is largely coterminous with the former Soviet Union).
Perhaps no religious tradition has been scrutinized for its compatibility with
democracy as has Islam—a 1,400-year-old faith practiced by almost two billion
people. More than 250 years ago, Montesquieu ([1748] 1900) famously argued that
“a moderate government is most agreeable to the Christian religion, and a des-
potic government to the Mohammedan.” Alexis de Tocqueville similarly thought
that Islam was inimical to democracy, arguing that Islam’s fusion of religion with
Introduction 11
politics meant that it could “never long predominate in a cultivated and democratic
age” (quoted in Hashemi 2009, 118). More recent testimonials to this supposed in-
compatibility of Islam and democracy were offered by the late English historian
Elie Kedourie (1994, 1), who wrote that “the idea of democracy is quite alien to
the mind-set of Islam,” and the American political scientist, Samuel Huntington
(1991, 307), who contended that, despite containing some pro-democratic elem-
ents, “Islamic concepts of politics differ from and contradict the premises of demo-
cratic politics.” Repeated cross-national studies by Fish (2002), Donno and Russett
(2004), and Teorell (2010) seem to validate this skepticism, finding a robust and
negative correlation between Islam and democracy, albeit one driven largely by
the durable autocracies of the Muslim world’s Arabic-speaking core (Stepan and
Robertson 2003).6 Although the persistence of Indonesian democracy since 1999
and Tunisia’s existence as a relatively stable democracy between 2011 and 2021
could be thought to constitute rejoinders to arguments about Islam’s incompat-
ibility with democracy, the decay of previously consolidated, Muslim-majority
democracies like Turkey and Mali give the hypothesis continued relevance.
A seventh contextual factor long thought to shape the possibility for demo-
cratic survival is the international environment confronting a given regime.
More than fifty years ago, Rustow (1970, 348) recognized the existence of “cases
where the major impulse to democratization came from the outside,” although
he left their analysis to scholars of international relations. Huntington (1991,
13) argued that “the occurrence and the timing of the third-wave transitions
to democracy” was shaped in part by “changes in the policies of external ac-
tors, most notably the European Community, the United States, and the Soviet
Union,” and the “demonstration effect of transitions earlier in the third wave in
stimulating and providing models for subsequent efforts at democratization.”
Brinks and Coppedge (2006, 464) point to the importance of superpower influ-
ence on the regime types of their clients, as well as to what they call “neighbor
emulation,” which they define as “a tendency for neighboring countries to con-
verge toward a shared level of democracy or nondemocracy.” In their study of
democratization in Latin America, Mainwaring and Pérez-Liñán testify to the
importance of neighbor influence, arguing that a “favorable regional polit-
ical environment, characterized by the existence of many democracies in Latin
America, increase[d]‌the likelihood of transitions from authoritarian rule to
competitive regimes and diminishe[d] the likelihood of breakdowns of existing
competitive regimes” (2013, 6). Observers of Eastern Europe, East Asia, and the
Middle East have noted the ways in which regional powers exert pressures on
their neighbors that lead to a kind of institutional isomorphism (Burnell 2010;
Coppedge et al. 2022; Darwich 2017; Lee 2018).
Table 1.1 contains the results of a time series, cross-sectional logistic regres-
sion analysis of the correlates of democratic breakdown for the eighty-three
Table 1.1. Regression Analysis of Correlates of Democratic Breakdown
Random effects logit Penalized logit
(1) (2) (3) (4) (5) (6)
GDP (logged) −0.744*** −0.891* −1.372*** −0.720*** −0.589 −0.981***

(0.207) (0.506) (0.420) (0.186) (0.373) (0.265)
GDP growth (%) −5.160* −4.045 −4.361 −5.194** −5.193 −5.982
(3.047) (4.254) (4.432) (2.533) (4.720) (4.671)
Share of neighboring −0.755 −6.911** −5.590* −0.704 −4.822** −3.583*
countries democratic (1.577) (3.179) (3.070) (1.427) (2.194) (2.062)
Years since last −0.00623 0.137* 0.116 −0.00725 0.0783* 0.0537

transition (0.0310) (0.0716) (0.0713) (0.0235) (0.0457) (0.0441)
Organization of −0.253 −1.967 −2.168* −0.206 −1.228* −1.332**
Islamic Cooperation (0.479) (1.214) (1.218) (0.423) (0.652) (0.637)
Arab League 2.266*** 4.246 5.242** 2.213*** 2.845* 3.875***
(0.822) (3.451) (2.524) (0.671) (1.474) (1.234)
Eastern Europe 0.928* 2.262** 1.821** 0.910** 1.756** 1.353**
(0.499) (0.997) (0.804) (0.441) (0.757) (0.607)
Ethnic −1.462 −1.209
fractionalization (1.103) (0.913)
index
State capacity −1.043* −0.890*
(0.548) (0.505)
Gini index 0.0309 0.0291
(0.0317) (0.0275)
Constant 2.492* 3.838 8.066** 2.373* 1.643 5.188***
(1.460) (4.447) (3.132) (1.301) (3.108) (1.788)
Observations 1,675 1,026 1,026 1,675 1,026 1,026
Number of clusters 82 72 72
(countries)
Note: Cells are coefficient estimates of regression models predicting the probability of democratic breakdown in
a given year. Models 1, 2, and 3 are random effects logits using Stata’s xtreg command, with standard errors clus-
tered by country. Models 4, 5, and 6 are penalized logits using the user-generated command firthlogit (Coveney
2021). Standard errors in parentheses.
* p < .1; ** p < .05; *** p < .01
countries that transitioned to democracy between 1974 and 2012.7 The purpose
of this exercise is to identify the factors most associated with democratic lon-
gevity and breakdown, and to subsequently calculate the breakdown risk for
each of the cases in this volume. The dependent variable, democratic breakdown,
Introduction 13
is a dichotomous variable that captures a shift from electoral democracy to au-

tocracy, measured up to 2017. Specifically, we used Mainwaring and Bizzarro’s
(2020) modified version of Lührmann et al.’s (2018) “Regimes of the World
Dataset,” which is based on the Varieties of Democracy (V-Dem) project.
Lührmann et al. consider a regime minimally democratic if (1) it meets min-
imum standards of free and fair elections; (2) it allows for some multiparty com-
petition; and (3) V-Dem’s electoral democracy score (which ranges between 0
and 1) is at least 0.50. Mainwaring and Bizzarro modified these coding rules to
exclude as either a transition or breakdown small, momentary increases above
or below the 0.50 electoral democracy threshold that were followed by a quick
reversion to the prior regime state. (For instance, a country whose democracy
score of 0.49 jumps in one year to 0.51 before settling back down at 0.49 would
not be considered to have undergone a transition followed by a breakdown.)
The key independent variables are:
• Economic development: This is measured as the natural log of the GDP per
capita (in 2010 constant US dollars), sourced from the World Bank’s World
Development Indicators.
• Economic inequality: This is the Gini index, which measures income ine-
quality on a scale from 0 to 1, where 1 corresponds to greatest inequality.
The source for this variable is the V-Dem dataset, which in turn draws
on data collected by the United Nations University’s World Institute for
Development Economics Research (Coppedge et al. 2018).
• Economic growth: This is the percentage change in GDP per capita for each
country year.
• Neighborhood effects: This is the share of countries in each geopolitical re-
gion (post-communist, Latin America, Middle East, sub-Saharan Africa,
Western Europe and North America, East Asia, Southeast Asia, South Asia,
the Pacific, and the Caribbean) that score 0.5 or above on the Varieties of
Democracy project’s electoral democracy index.
• Years democratic: Calculated as the number of years since the last demo-
cratic transition.
• Ethnic fractionalization: This is the probability that any two randomly
selected individuals from a given country in a given year will be from dif-
ferent ethnic groups, calculated by Drazanova (2019) from data gathered by
the Composition of Religious and Ethnic Groups project at the University
of Illinois at Urbana-Champaign. The index ranges (theoretically) from 0,
indicating that every citizen is from the same ethnic group, to 1, where there
are as many ethnic groups as individuals.
• State capacity: This is a measure of state infrastructural power devised by
Hanson and Sigman (2013) that captures three dimensions of state capacity
(extractive, administrative, and coercive). Values range from –3 to 3.
Finally, cultural and religious influences are captured with dummy variables
for membership in the Organization of Islamic Cooperation (Albania,
Bangladesh, Benin, Comoros, Guyana, Indonesia, Iraq, Lebanon, Libya,
Maldives, Mali, Niger, Nigeria, Senegal, Sierra Leone, Suriname, Turkey,
and Tunisia), the Arab League (Comoros, Iraq, Lebanon, Libya, Mauritania,
and Tunisia), and for being situated in Eastern Europe (Albania, Armenia,
Belarus, Bulgaria, Croatia, Czech Republic, Estonia, Georgia, Hungary, Latvia,
Lithuania, Macedonia, Moldova, Poland, Romania, Russia, Serbia, Slovakia,
Slovenia, and Ukraine). Summary statistics for each of these variables are
presented in Table A1.1 in the appendix.
Due to missingness in the data for ethnic fractionalization, state capacity,
and economic inequality, column 1 of Table 1.1 presents the results of a random
effects logit without those variables included. Column 2 presents the regression
results with all variables included, at the expense of losing nearly 40 percent of
observations. In order to determine whether the differences in the sign, signifi-
cance, and magnitudes of the coefficients in columns 1 and 2 are due to the inclu-
sion of the additional variables in Model 2 or to the loss of observations, column
3 reports the results of a regression with only the smaller number variables in
column 1 and only the smaller number of cases in column 2. Columns 4, 5, and
6 repeat the analyses in columns 1, 2, and 3 using a penalized logit estimator to
account for the rarity of democratic breakdowns in the data (forty-five in 1,725
regime-years). The results of these latter regressions are substantially similar
to the regressions without rare-events adjustments, although effect sizes and
standard errors are generally smaller.
As expected, the relationships between democratic breakdown and GDP, GDP
growth, regional regime characteristics, and state capacity are negative in most of
the models—meaning that democratic breakdown is less likely in countries that
are richer, growing economically, situated among other democracies, and which
possess high state capacity. However, none of these variables is statistically signif-
icant across all six models reported in Table 1.1, which testifies to the uncertain
relationship between structural factors and democratic survival.
Figures 1.3a through 1.3g plot the marginal effects of our key continuous
predictors (GDP per capita, GDP growth, regional share of democracies, years
democratic, ethnic fractionalization, state capacity, and income inequality) on
the probability of democratic breakdown.8 Figure 1.3a shows that, holding all else
equal, a country whose logged GDP per capita is in the 75th percentile (approxi-
mately USD 8,500) has a 1.2 percent chance of experiencing a democratic break-
down in a given year, which is about a quarter of the probability that democracy
will break down in a country whose GDP is in the 25th percentile (approximately
USD 1,500). Figure 1.3b displays the effect of economic growth on the probability
of democratic breakdown. A country in the 75th percentile of economic growth
(a) GDP (b) GDP growth (c) Share of neighboring countries democratic
.6 .08
.2
.4 .06
.15
Pr(Breakdown)
Pr(Breakdown)
Pr(Breakdown)
.04
.1 .2
.02
.05 0
0
0 −.2
6 7 8 9 10 −.4 −.2 0 .2 .4 .6 0 .2 .4 .6 .8
GDP per capita, logged % GDP growth Share of neighboring countries that are democratic
(d) Years since last transition (e) Ethnic fractionalization (f) State capacity
.1
.06 .4
.08
.3
Pr(Breakdown)
Pr(Breakdown)
Pr(Breakdown)
.04
.06 .2
.02
.04 .1
0
.02 0
−.02 0 −.1
0 10 20 30 40 −2 −1 0 1 2
Years democratic Ethnic fractionalization State capacity
(g) Inequality
.15
.1
Pr(Breakdown)
.05
−.05
20 40 60 80
Gini index
Figure 1.3. Marginal effects of continuous predictors

Sources: For (a) and (b), World Development Indicators, n.d.; for (c), (d), and (g), Coppedge et al. 2018; for (e), Drazanova 2019; and for (f), Hanson and Sigman 2013.
experiences a probability of breakdown (0.28 percent) that is ten times as small

as the probability of breakdown for a country that is in the 25th growth percen-
tile. The effect of neighbor emulation, displayed in Figure 1.3c, is relatively flat
(which is to be expected, given that the coefficient on this variable is insignificant
in Model 1). Based on the simulation for Model 1, a democracy in a neighbor-
hood that is 55 percent democratic (corresponding to the 75th percentile of re-
gions on this measure) has, all else equal, a 2.5 percent chance of breaking down,
whereas one in a neighborhood that is 45 percent democratic (corresponding to
the 25th percentile), has a 2.7 percent chance of breaking down. State capacity is,
as expected, negatively associated with breakdown, as can be seen in Figure 1.3f.
A democracy with around the 25th percentile of state capacity has a 3.2 percent
chance of breaking down in a given year, while one in the 75th percentile of state
capacity has a chance of breaking down of around 1 percent.
The effect of the years since democratic transition is sensitive to specification,
with sign and significance changing depending on which variables are included.
The Gini index and ethnic fractionalization do not achieve significance.9
The effects of the dichotomous regional predictors on democratic breakdown
are presented in Figures 1.4a through 1.4c. Here, the results offer mixed sup-
port for the conventional wisdom. Membership in the Organization of Islamic
Cooperation, which is our rough proxy for a country’s Muslim majority status,
is negatively associated with democratic breakdown—significantly so in Model
5 when controls for state capacity, economic inequality, and ethnic fractionali-
zation are included. However, the indicator for membership in the Arab League
is a positive and almost always significant correlate of democratic breakdown
(Model 2 is the only exception). A non-Arab democracy has a 2.4 percent chance
of breakdown in a given year, while a democracy that is also an Arab League
member has a 17 percent chance of breaking down in a given year. Being situ-
ated in Eastern Europe is consistently a risk factor for democratic breakdown, al-
though the substantive impact is smaller than being an Arab League member: the
average Eastern European democracy has a 5.4 percent chance of breaking down
in a given year, while the average non–Eastern European democracy’s chance of
breaking down is around 2.3 percent.
Overall, the results reported in Table 1.1 provide mixed support for long-
theorized arguments about the relationship between contextual factors and
countries’ democratic prospects. Moreover, the instability in the magnitudes of
the coefficients and their significance indicate the need for caution about strong
claims about the influence of these contextual variables. The simple fact of the
matter is that these variables cannot explain all of the observed variation in dem-
ocratic breakdown. Thus, although the chapters that follow all clearly situate the
case studies in the context of these macro variables, they also emphasize how
actors have maneuvered around them to keep democracy alive despite the odds.
(a) Arab League
.4
.3
Pr(Breakdown)
.2
.1
0 1
Arab League
(b) Organization of Islamic Cooperation

.04
Pr(Breakdown)
.03
.02
.01
0 1
OIC membership
(c) Eastern Europe

.1
.08
Pr(Breakdown)
.06
.04
.02
0 1
Eastern Europe
Figure 1.4. Marginal effects of qualitative predictors

Note: The Organization of Islamic Cooperation includes Albania, Bangladesh, Benin, Comoros,
Guyana, Indonesia, Iraq, Lebanon, Libya, Maldives, Mali, Niger, Nigeria, Senegal, Sierra Leone,
Suriname, Turkey, and Tunisia; the Arab League includes Comoros, Iraq, Lebanon, Libya,
Mauritania, and Tunisia; and Eastern European countries are Albania, Armenia, Belarus, Bulgaria,
Croatia, Czech Republic, Estonia, Georgia, Hungary, Latvia, Lithuania, Macedonia, Moldova,
Poland, Romania, Russia, Serbia, Slovakia, Slovenia, and Ukraine.
The Cases
We have identified several factors that scholars have traditionally deemed im-
portant to a country’s chances of getting and keeping democracy: economic
development; the absence of stark economic inequality; economic growth (or,
at the very least, the absence of a prolonged economic crisis that leads to mass
disillusionment with democratic government); ethnic homogeneity and the
absence of deep identity cleavages; strong states capable of maintaining order
and governing effectively; mass cultures compatible with democratic norms
of tolerance and pluralism; and neighboring governments and foreign patrons
that are themselves democratic and supportive of democratic rule. Although
these factors have some explanatory power, they are nonetheless collectively
unable to account for the durability of the many democracies around the
world that are poor, ethnically fragmented, with traditional cultures, situated
amid a sea of autocracies, and with troubled economies that are often stagnant
or in crisis.
To explore how democracies survive under such conditions, the book analyzes
the regime trajectories of nine countries that experienced considerable (if varying)
democratic spells despite facing great challenges: Argentina, Benin, Georgia,
India, Indonesia, Moldova, South Africa, Timor-Leste, and Ukraine. Two criteria
guided our selection of cases. The first is that a country should have survived as a
democracy for a long period of time (as of 2017, when the project was launched).
Using the data compiled by Bizzarro and Mainwaring (2020), we calculated that
the average duration of a third-wave democracy (as of 2017) was 12.3 years. We
therefore established this duration as the initial cutoff for inclusion as a core case in
the volume. Of the nine cases in this volume, seven meet this criterion: Argentina,
Benin, Georgia, India, Indonesia, Timor-Leste, and South Africa. Two additional
cases, Moldova and Ukraine, have oscillated between democratic and autocratic
spells and do not qualify as long-lasting democracies, but they are among the most
democratic of the twelve former Soviet Republics outside of the Baltics.
The second criterion for case selection was that the cases be deemed improb-
able by the lights of structural accounts of democratic survival. In all of these
countries—save possibly for Argentina—the factors that facilitate democratic
survival were largely absent. Most experienced low levels of economic devel-
opment and educational attainment. Many inherited enfeebled administrative
apparatuses and coup-prone militaries. Some confronted repeated economic
crises. And some were located in parts of the world in which democracies were
few, or were embedded in cultures and religions very different from those in
which modern democracy was incubated.
Table 1.2 presents the average score for each country on each of the predictors
analyzed in Table 1.1, for the entirety of the period for which it appears in the
Table 1.2. Values for Each Case on Key Independent Variables (Socioeconomic Indicators Averaged over Democratic Lifespan)
Country GDP per GDP Regional Years Ethnic State Gini Islamic Arab Eastern
capita growth (%, democracy democratica fractionalization capacity European
geometric share (%)
mean)
Argentina 8,511.71 1.01 55.74 34 0.13 0.20 45.88 0 0 0

Benin 726.89 2.04 41.22 26 0.77 −0.93 36.91 1 0 0
Timor-Leste 3,087.56 4.18 50.79 16 0.83 −1.42 32.5 0 0 0
Georgia 3,515.77 5.67 54.39 13 0.38 0.38 40.77 0 0 1
India 879.18 3.96 38.36 41 MISSING −0.03 33.5 0 0 0
Indonesia 3,006.01 3.89 49.32 18 0.8 0.13 35.3 1 0 0
Moldova 1,713.23 3.29 52.16 19 0.48 −0.31 34.77 0 0 0
South Africa 6,797.02 1.31 43.81 22 0.86 0.72 59.75 0 0 1
Ukraine 2,665.59 1.69 52.97 12 0.39 −0.16 29.91 0 0 1
Sources: Coppedge et al. 2018; Drazanova 2019; Hanson and Sigman 2013; World Development Indicators, n.d.
a According to coding scheme in Mainwaring and Bizzarro (2020).
.25
.2
.15
Pr(Breakdown)
.1
.05
1978 1982 1987 1992 1997 2002 2007 2012 2017
India Spain
Figure 1.5. India (1978–2017), probability of democratic breakdown

(benchmarked against Spain)
data. Table 1.2 highlights some of the daunting challenges the countries in this
volume faced in building long-lasting democracies. Most of the countries are
poor—Benin and India strikingly so. Argentina, South Africa, and Ukraine have
had poor economic performance as measured by per capita GDP growth. India,
Benin, and South Africa are situated in regions dominated by authoritarian
regimes. Benin, Timor-Leste, India, and South Africa have great ethnic and lin-
guistic diversity. Benin and Timor-Leste have shaky state capacity. And South
Africa features some of the starkest income inequalities on the planet.
Figures 1.5 through 1.13 display the breakdown probabilities—drawn from
Model 1 in Table 1.1—of each of the cases in this volume. In order to enable the
reader to more easily comprehend the outsized magnitude of these (objectively
small) breakdown probabilities in a given year, we include in each graph the
breakdown probabilities for Spain—one of the first third-wave democracies—
over the life of its democratic regime. According to these simulations, among our
six cases of long-lasting democracy, India in the years just after redemocratization
in 1977 faced particularly long odds. Moldova and Ukraine also had years in
which the predicted odds of breakdown were vastly higher than those for Spain.
For all of these cases, the probability of breakdown in any particular year was not
startlingly high—although it was much higher than Spain’s—but the cumulative
probability of survival over the life span of these democracies ranges from poor
to moderate.
Introduction 21
The first case in the book is India, which is perhaps the quintessential democ-
racy in a hard place, having long been identified by scholars as a stubbornly per-
sistent rejoinder to most theories of democratization. (The annual breakdown
probabilities for India, based on our model, are presented in Figure 1.5.) Poor,
ethnically heterogeneous, with a caste system that seems incompatible with
democracy’s assumptions of formal political equality, and located in a neighbor-
hood with precious few democracies to speak of, India seems almost entirely be-
reft of the things that scholars believe to be conducive to democracy. And yet,
as contributor Ashutosh Varshney points out, India has been a democracy for
seventy-two of the last seventy-four years. Although India has always fallen short
of the liberal ideal and currently faces a populist challenge not unlike those faced
by some other established democracies, it is nonetheless perhaps democracy’s
most remarkable success story.
Varshney’s explanation for how India has eluded the democratic collapse
predicted for it by most theories of democratization emphasizes the importance
of the beliefs and values of that country’s modern founders, and, in particular, its
first prime minister, Jawaharlal Nehru. Possessed of a genuine belief in popular
sovereignty, and in the equality of India’s people regardless of religion, caste, or
class, Nehru and his allies wrote a constitution that guaranteed the right to po-
litical organization and emphasized the inviolability of democratic procedure.
But although Varshney attributes the birth of Indian democracy to the values
of the country’s founders, he attributes democratic persistence to the fact that
India’s political elites have come to see democracy as being in their interests.
In particular, the proliferation of political parties at all levels of the country has
ensured that “there is enough countervailing power available in the system to
oppose the violation” of democratic norms and procedures. Thus, when Indira
Gandhi suspended democracy in 1975 on the pretense of a national emergency,
she was only able to hold the line for nineteen months, eventually giving in to
pressure for new elections. Once democratic political life was resumed, Gandhi
was voted out, and Indian political elites amended the constitution to make fu-
ture ant-idemocratic trespasses difficult to repeat. Courts and an independent
election commission have helped keep democracy intact, and the dispersion of
power across India’s states and union territories has given many parties a stake
in preserving the regime. The story of India, then, is one of founders motivated
by a commitment to democracy who created a system that eventually gave rise
to a highly pluralistic federal landscape in which democracy was, until recently,
by far the dominant game in town. Varshney notes, however, that the rise of
the distinctly illiberal, Hindu nationalist Bharatiya Janata Party (BJP) has intro-
duced new threats to Indian democracy, and serves as a dramatic illustration of
the extent to which democracy’s survival hinges on political parties’ normative
commitments.
.04
.03
Pr(Breakdown)
.02
.01
1978 1982 1987 1992 1997 2002 2007 2012 2017
Indonesia Spain
Figure 1.6. Indonesia (2000–2017), probability of democratic breakdown

Indonesia, the world’s third most populous democracy after India and the
United States, also faced daunting challenges (see Figure 1.6 for Indonesia’s
predicted probabilities of democratic breakdown). That country’s per capita GDP
of USD 4,284 may be approximately double that of India, but it remains in the
bottom half of countries. Its population of 267 million is about one fifth of India’s,
but Indonesians speak more than seven hundred languages, live on one thousand
islands scattered across the world’s largest archipelago, and mostly practice Islam,
a faith that some cross-national studies of regime type find to be negatively associ-
ated with democracy. On top of all of this, Indonesia’s democracy, inaugurated in
1998, came after thirty-two years of a brutal, military-backed dictatorship that had
been baptized in the blood of more than half a million Indonesian communists.
Given these characteristics, argues Dan Slater, Indonesia is vulnerable to
four modes of democratic collapse: a failure of the state, a military takeover, a
transition to electoral authoritarianism, or the rise of a majoritarian, illiberal
“democracy” in which minority rights are trampled. That Indonesia has
managed to avoid these pitfalls is, he argues, a function of the legacies of the
pre-authoritarian period. If the story of India is one of democratic legacies
bequeathed by committed democrats, which could only briefly be undone,
the story of Indonesia is one of autocrats who unwittingly erect the scaffolding
that later helps to hold up (an admittedly imperfect) democratic edifice. Slater
emphasizes four inheritances, three from the authoritarian period known as the
New Order (1966–98). The first is an encompassing national identity that helped
Introduction 23
to instill a sense of belonging among the country’s diverse peoples and to stave
off ethnic polarization and secessionism. The second is the emergence of an in-
dependent, technocratic bureaucracy that would prove capable of governing
even after the removal of the dictator, thus precluding the kind of administrative
nonperformance that routinely sours citizens on new democracies. The third
was an autocratic “ruling party” that remained intact through the transition and
which kept former regime satraps invested in the new democracy by providing
them with a channel for participating in it. And finally, the autocrat’s strategy of
dividing and ruling over the military apparatus left it without the ability to act
collectively to abrogate democracy even if it had wanted to.
Slater’s emphasis on how authoritarianism can structure the political land-
scape in ways that are later conducive to democratic survival offers a counter-
point to Varshney’s emphasis on the values and beliefs of political leaders at
founding moments. The differing emphases of these two contributions lay bare a
key point of contention that comes up time and again in this book’s narratives of
democratic survival: Though most accounts in this book agree that actors either
keep democracy alive or kill it, they differ in the extent to which they believe that
their democracy-sustaining (or democracy-destroying) behaviors are the result
of beliefs versus a narrowly instrumental calculation of costs and benefits.
In her analysis of democracy’s emergence and survival in Benin and South
Africa, contributor Rachel Riedl takes the latter position. Both countries are
hard places for democracy, as evidenced by the annual breakdown probabilities
.15
.1
Pr(Breakdown)
.05
1978 1982 1987 1992 1997 2002 2007 2012 2017
Benin Spain
Figure 1.7. Benin (1992–2017), probability of democratic breakdown

.03
.02
Pr(Breakdown)
.01
1978 1982 1987 1992 1997 2002 2007 2012 2017
South Africa Spain
Figure 1.8. South Africa (1996–2017), probability of democratic breakdown

displayed in Figures 1.7 and 1.8. Benin (Figure 1.7) is impoverished (with a cur-
rent GDP per capita of USD [2010] 1,200), ethnically and religiously diverse, and
plagued by low state capacity. South Africa (Figure 1.8) is more prosperous and
has a more modern, effective state, but it grapples with extreme economic ine-
quality stemming from a long history of white domination of the country’s black
majority and other non-white minorities. Both countries are located in a part of
the world that has few long-lived democracies, further complicating the likeli-
hood of getting and keeping democratic government.
For Riedl, the emergence and survival of democracy in these two hard places
was not a function of ideological commitments to democratic government on
the part of political leaders. In both places, if incumbent elites had been able to
maintain autocracy, or if rising challengers had proven able to erect autocra-
cies of their own, there is reason to expect they would have. But in both places,
Riedl argues that democracy emerged and was sustained as the result of pacts by
incumbents who understood that the authoritarian status quo was unsustainable
and opponents who realized that economic and social stability required com-
promise with the ancien régime. As in Indonesia, democracy survived in South
Africa and Benin because the old elite was given a place in the new political
system and thus disincentivized from engaging in subversion.
Chapter 5, by Lucan Way, explores the fate of democracy in three of the former
Soviet Republics: Georgia, Ukraine, and Moldova (breakdown probabilities
.1
.08
Pr(Breakdown)
.06
.04
.02
1978 1982 1987 1992 1997 2002 2007 2012 2017
Georgia Spain
Figure 1.9. Georgia (2005–2017), probability of democratic breakdown

.2
.15
Pr(Breakdown)
.1
.05
1978 1982 1987 1992 1997 2002 2007 2012 2017
Ukraine Spain
Figure 1.10. Ukraine (1995–1997, 2007–2013), probability of democratic

breakdown (benchmarked against Spain)
.25
.2
Pr(Breakdown)
.15
.1
.05
1978 1982 1987 1992 1997 2002 2007 2012 2017
Moldova Spain
Figure 1.11. Moldova (1995–2004, 2010–2017), probability of democratic

breakdown (benchmarked against Spain)
displayed in Figures 1.9, 1.10 and 1.11, respectively). According to Way, these cases
are instances not so much of democratic survival amid challenge as of democratic
“moments” embedded within otherwise troubled regime trajectories. Each case
counts as a “hard” one by the lights of this book. Even the case with arguably the
best democratic record, Georgia, is a hard place on a number of dimensions: its per
capita GDP of less than USD 5,000 places it in the bottom half of countries; 83 per-
cent of the population is Orthodox Christian and another 11 percent are Muslims,
two religions that scholars have identified as being inconducive to democracy; and
around 13 percent of the country is made up of ethnic minorities, including Azeris
(6 percent), Armenians (4 percent), and Russians (1.2 percent) (CIA 2021).
In Way’s account, the democratic moments experienced by these three cases
are the results of three factors that act to constrain would-be autocrats: The first
is the relative weakness of the state apparatus, which results in what he calls “plu-
ralism by default.” The second is the inheritance from the authoritarian period of
political parties and independent media organizations that act to check the power
of new regimes. And the third is the looming specter of Russian influence, which
has rendered some of the region’s anti-Russian political leaders—who might have
indulged their authoritarian tendencies—receptive to external pressures to min-
imally uphold democracy and the rule of law in order to maintain US support.
One of the strengths of the accounts that attribute democratic survival to
the constraints faced by political elites is that the origins of those constraints
are relatively transparent, generally residing in conditions inherited from the
Introduction 27
authoritarian past. (Often less clear in these accounts is why the foundational
democratic pacts persist over time, when they do.) In contrast, those who em-
phasize the importance of normative commitments to democracy often do not
explain where those commitments came from. The final two chapters in this
volume take this task head-on, offering accounts of democratic survival that
pay close attention to the processes that convert political leaders into committed
democrats.
The first of the chapters explains how democracy took root and survived in
Timor-Leste, a small island nation that was occupied successively by Portugal
(1702–1975) and Indonesia (1975–99), against which it fought a decades-long
war of independence that claimed more than one hundred thousand lives. Of
all of the cases in this volume, Timor-Leste is one of the hardest—in addition to
its long history of internal war, it remains a poor country with a very weak state
and a high degree of ethnolinguistic diversity. Timor-Leste’s relatively high prob-
abilities of democratic breakdown since the onset of its democracy in 2002 are
displayed in Figure 1.12.
Nancy Bermeo argues that Timor-Leste’s democracy has survived these
challenges as a result of changes to the polity and its leaders that took place over
the course of its long war for independence from Indonesia. In the waning years
of the Portuguese colonial period, Timor-Leste’s political and economic elite
were deeply divided over their identities and their preferences for the country’s
future. Some wanted independence, others wanted union with Indonesia, still
.2
.15
Pr(Breakdown)
.1
.05
–.05
1978 1982 1987 1992 1997 2002 2007 2012 2017
Timor-Leste Spain
Figure 1.12. Timor-Leste (2002–2017), probability of democratic breakdown

others sought union with Australia. Some were communists, others monarchists,
still others Christian conservatives. These cleavages meant that Timor’s political
elites were unable to erect an independent government in the wake of Portugal’s
withdrawal in 1975, and their resulting infighting opened the door for Indonesia’s
invasion shortly thereafter.
However, according to Bermeo, the Indonesian occupation and the war to end
it set in motion several processes that ultimately endowed Timor-Leste with the
ingredients for its present democracy. First, the country’s anti-democratic political
and economic elites were so brutalized by the occupation that they either disappeared
from the scene or gave themselves over to a new, democratic mindset. Second, the
independence struggle forged a unified and inclusive national identity among the
people of Timor-Leste, enabling them to overcome the ethnic and ideological
cleavages that had contributed to their prior vulnerability to Indonesian aggression.
Third, the war gave rise to a number of political leaders with good reputations and
large constituencies who could convert their renown into votes, resulting in a post-
independence political landscape that was pluralistic rather than dominated by a
single group. Fourth, and finally, the exigencies of the national struggle had forced
the militias to coalesce into a national army divorced from any one political party and
subordinate to the authority of democratically elected leaders.
If Timor-Leste is one of the hardest places for democracy explored in this
volume, the final case study, of Argentina, might appear at first glance to be the
easiest. With the highest per capita GDP among our cases, a relatively capable
state, limited ethnic fragmentation, and, by the 1990s, a mostly favorable neigh-
borhood, that country would appear to be free of some of the challenges that
confront the other countries in this volume. And yet, Scott Mainwaring and
Emilia Simison point out, Argentina’s democracy suffered five breakdowns in
the twentieth century. In the years since democracy was re-established in 1983,
it confronted three punishing economic crises, a dramatic increase in inequality
and poverty, several military rebellions, and a tumultuous period in the close of
2001 in which the country had five presidents in thirteen days. These economic
difficulties are reflected in the country’s elevated probability of democratic
breakdown (compared to Spain) since its transition in 1983 (see Figure 1.13).
For those worried about democratic collapse in established democracies of the
Western world, the case of Argentina is therefore instructive.
How did Argentina overcome its authoritarian past and hang onto democracy
when other countries that have faced much lesser economic crises experienced
economic breakdowns? Mainwaring and Simison emphasize the importance
of normative commitments to democracy on the part of the country’s political
parties, labor and social movements, and business leaders. They also empha-
size the policy moderation of all key actors—a profound contrast to the situa-
tion during the country’s previous democratic experience of 1973–76. And like
Bermeo, they identify the source of these normative commitments and the shift
Introduction 29
.04
.03
Pr(Breakdown)
.02
.01
1978 1982 1987 1992 1997 2002 2007 2012 2017
Argentina Spain
Figure 1.13. Argentina (1984–2017), probability of democratic breakdown

to policy moderation: Argentina’s experience with a brutal and inept military

dictatorship from 1976 to 1983. According to Mainwaring and Simison, those
years were so catastrophic for Argentina, marked by “torture, ‘disappearances,’ a
reckless and failed war,” and repression of labor unions and political parties, that
political leaders across the political spectrum came to value democracy. They
acted to uphold the democratic system even during severe economic downturns,
hyperinflation, sharp increases in inequality, and dramatic increases in poverty
and unemployment. Thus, when groups of officers engaged in military rebellions
during a particularly difficult period for the Argentine economy, masses and
elites were unified in rejecting them. As Guillermo O’Donnell (1986, 15) wrote
thirty-six years ago when assessing the prospects of Latin American democra-
cies that had recently emerged out of brutal autocratic regimes, “Largely as a con-
sequence of the painful learning induced by the failures of those regimes and
their unprecedented repression and violence, most political and cultural forces
of any weight now attribute high intrinsic value to the achievement and consoli-
dation of political democracy.”
The Importance of Normative Commitments
Each of these chapters demonstrates powerfully that actors matter, but not in the
manner assumed by naïve voluntaristic accounts. It is not the case that democ-
racy is made and kept by great leaders bound only by their skills and passions.
Democratic survival is a process marked by constraint. In each chapter, we see

that actors who could have abrogated or weakened democracy were prevented
from doing so—sometimes by “external” constraints in the form of institutional
shackles or the countervailing power of competing actors, but sometimes by “in-
ternal” constraints, in the form of deeply ingrained normative commitments. In
the volume’s concluding chapter, Scott Mainwaring explores the relative impor-
tance of these two forms of constraint for understanding democracy in hard places.
Although Mainwaring argues that both internal and external constraints are
at work in most cases, he nonetheless makes a powerful argument for greater
attention to the former. Formal institutions such as courts and legislatures do
sometimes constrain rulers, helping keep democracy intact against executive
encroachments, but these institutions tend to be weak in democracies in hard
places; presidents and prime ministers can often control, manipulate, or bypass
them. A balance of power among partisan actors is helpful to democratic survival,
forcing actors to accept a democratic compromise. But the balances of power
that lead to the establishment of democracy usually do not persist. In the cases
in this volume, the African National Congress has enjoyed unassailable electoral
hegemony since 1994. In Benin, Georgia, India, and South Africa, parties some-
times won landslide electoral victories. They had the opportunity to steamroll the
opposition, and yet, they refrained from doing so (until 2018 in Benin).
In short, the external constraints highlighted by the literature are likely nei-
ther necessary nor sufficient for democratic survival. The preferences of actors
over regime type matter. Some actors value the rights and procedures that are
defining features of democracy. These clearly help to sustain democracy. At the
other end of the spectrum are committed authoritarians who believe that dif-
ferent forms of dictatorship—communism, fascism, theocracy, populist illiberal
democracy—are the normative ideal. These actors weaken democracy. Likewise,
radical actors who demand wholesale revisions of the status quo—or who defend
the status quo at all costs—raise the stakes of a democratic game and contribute
to the possibility of democratic breakdown.
Surveying the cases in the volume, Mainwaring finds evidence for the im-
portance of normative commitments to democracy across the cases. Using
data from the Varieties of Party Identity and Organization (V-Party) project, he
shows that none of the cases of democratic survival in our sample featured highly
illiberal political parties for the period when they were undeniably democratic.
The emergence of highly illiberal parties (such as the BJP in India) do not so
much constitute a rejoinder to the notion that normative commitments matter as
a clear warning of heightened risk of democratic breakdown (as seen in Benin in
2019, Moldova in 2001 and 2009, and Ukraine in 2010).
But if normative commitments matter, where do they come from?
According to Mainwaring, they do not simply emerge from thin air or the
Introduction 31
biographical details of particular leaders. The collective actors who can make or
break democracy—and political parties in particular—develop programmatic
commitments to regime type that they inculcate into members and militants and
which often prove durable over time—just as other programmatic commitments
do (Berman 1998). In the case of Argentine parties, this commitment to democ-
racy emerged as part of a process of repudiating an authoritarian past. In the case
of India’s Congress Party, the commitment to democracy and liberal ideals was
shaped by an anti-colonial struggle that had at its center the liberal ideals of indi-
vidual rights and freedom from domination. In Timor-Leste, the commitment to
democracy took shape under the shadow of Indonesian occupation and a bloody
civil war. There is every reason to expect commitments forged in this manner to
endure. Though we would not go as far as Berman (1998, 207) to say that idea-
tional motivations are so powerful as to render “political actors . . . insensitive
to changes in their environment and relatively unconcerned with ‘cost-benefit’
calculations,” we do believe that collective actors’ normative commitments to de-
mocracy, once formed, constitute important “internal” constraints on their be-
havior. In some instances, these internal constraints can prove as important as
the counterbalancing power of competitors or the shackles of institutions.10
Conclusion
We opened this book with testimonials to the peril in which democracy finds itself
around the world, including—and perhaps especially—in wealthy countries where
it was once thought impregnable. The possibility of democratic collapse where we
least expected it has added new urgency to the age-old inquiry into how democracy,
once attained, can be made to last. This book argues that scholars and practitioners
interested in this question can learn much from democratic survivals that were as
unexpected as the democratic erosions currently feared in some corners of the de-
veloped world. Just as social scientists long believed that well-established, Western,
educated, industrialized, and rich democracies were immortal, so too did they as-
sign little chance of democracy to countries that lacked these characteristics. And
yet, in defiance of decades of social science wisdom, over the course of the last half-
century, many countries that were bereft of the hypothesized enabling conditions
for democracy not only got it but kept it, year after year after year. What is the se-
cret of democratic longevity in such hard places? What has enabled countries such
as India, Indonesia, and South Africa to hang onto democratic government de-
spite poverty, inequality, and ethnic fragmentation? How did Timor-Leste forge
a durable democracy out of the ruins of civil war? How did Argentina manage to
overcome an authoritarian past and keep democracy alive despite repeated eco-
nomic depressions? This book brings these and other unlikely cases of democratic
survival front and center in an effort to derive lessons about what makes democ-
racy stick, especially during moments of tumult and crisis.
Structure is not a prison. As the cases in this book demonstrate, Rustow (1970)
was correct when he observed that there are no background conditions whose
presence is absolutely necessary for democracy to emerge and survive. Instead,
democracy is made and upheld by collective political actors—parties, executives,
movements, militaries—that must make difficult choices structured not just by
country-level factors and international contexts, but also by their own norma-
tive commitments. Presidents facing electoral defeat may abide by the results or
use their control over the instruments of power to distort or ignore them entirely.
Military officers chafing under economic hardship can either respect incumbents
or throw them out. Opposition parties observing mass discontent with a regime can
either wait until the next election or go “knocking on the barracks door” (Stepan
1988, 128). The latter outcomes are more likely in “hard” places than in “easy” ones,
but, as the cases in this volume demonstrate, they are not predetermined. If the
accounts that follow demonstrate that democratic potential exists in places that
dominant structural theories would consign to authoritarianism, so too do they
demonstrate that nowhere should its continued existence be taken for granted.
If this volume is to make any contribution to the global fight to preserve democ-
racy, it is to emphasize for readers that democracy is nowhere assured and nowhere
doomed, but rather lives or dies depending on what political actors believe and do.
Notes
1. Diamond, Larry. 2019. “The Global Crisis of Democracy.” The Wall Street Journal, May
17, 2019, https://www.wsj.com/articles/the-global-crisis-of-democracy-11558105463.
2. In a later contribution, Przeworski (2005, 266–67) finds that the prospects of future
electoral victory “are neither sufficient nor necessary for democracy to survive.” In
poor countries, he argues, the income gain from ruling as a dictator might outweigh
the income gain to be had from winning democratic elections, while the cost of being
a loser in both circumstances is the same. In rich countries, in contrast, it may be
better to be a perpetual loser under a democratic system than to risk being a loser
under an authoritarian one.
3. Przeworski (2005, 265) admits that actors may have a “preference for democracy,
independently of income,” but he argues that this would only operate at high levels
of income (i.e., not in “hard places). He writes, “as the marginal utility of consump-
tion declines, the preference for democracy (or against dictatorship) overwhelms the
eventual consumption gain from becoming a dictator.”
4. An important exception to this widely shared finding is Acemoglu et al. (2008).
5. As Ansell and Samuels (2010, 1544) note, Boix (2003) and Acemoglu and Robinson
(2006) offer different predictions about the level of income inequality that is most
conducive to democratization, but the logic of their arguments is similar.
Introduction 33
6. Andrew March (2011) offers an alternative perspective on Islam’s compatibility with

democracy, and particularly what he calls “liberal citizenship.”
7. For exact details of coding democratic transitions and breakdowns, see Mainwaring
and Bizzarro (2020, 1561, note 3).
8. All marginal effects are calculated using the results in column 1 of Table 1.1, except
for the marginal effects of ethnic fractionalization, the Gini index, and state capacity,
which are calculated based on the results in column 2.
9. Although non-significant, the negative correlation between ethnic diversity and
democratic breakdown runs contrary to expectations. It is premature to speculate
about the cause of a correlation that has yet to be consistently demonstrated, but it
could suggest that social diversity renders it harder for single parties or groups to
monopolize power and transform competitive regimes into autocratic ones. This
does not mean that the conventional wisdom—which sees ethnic diversity as a chal-
lenge to democracy—is wrong, only that the difficulty that diversity poses could be in
getting democracy, and not in keeping it.
10. In their book, Costly Democracy: Peacebuilding and Democratization After War,
Zurcher et al. (2013) find that the success of democratization after civil wars hinges
in part on the extent to which political actors “demand” democracy, which can stem
from a combination of self-interest and normative belief.
Appendix
Table A1.1 Summary Statistics
Variable Obs. Mean Std. Min. Max.

Dev.
Breakdown 1,731 0.026 0.159 0 1

GDP (logged) 1,682 8.249 1.135 5.777 10.383
GDP growth 1,675 0.023 0.046 −0.407 0.599
Share of neighboring countries 1,731 0.513 0.127 0.103 0.794
democratic
Years since last transition 1,731 12.270 8.953 0 42
Organization of Islamic Cooperation 1,731 0.162 0.369 0 1
Arab League 1,731 0.020 0.141 0 1
Eastern Europe 1,731 0.224 0.417 0 1
Ethnic fractionalization index 1,588 0.436 0.253 0.003 0.889
State capacity 1,118 0.209 0.791 −2.377 2.015
Gini index 1,656 41.567 10.267 21.2 74.3
2
India’s Democratic Longevity and
Its Troubled Trajectory
Ashutosh Varshney
At the core of this chapter lies a paradox.* On one hand, India is the longest
lasting democracy of the developing world; on the other hand, since 2014 a dem-
ocratic decline has unquestionably set in. In their recent reports, the two most
widely read annual assessments of democracy worldwide, by Freedom House
and the V-Dem Institute, have noted India’s democratic retrogression in no un-
certain terms. Freedom House now calls India only “partly free,” and the V-Dem
Institute says India has become an “electoral autocracy” (Freedom House 2021b;
V-Dem Institute 2021). Whether or not we find these terms precise, India’s dem-
ocratic diminution is not in doubt. With the rise of Narendra Modi to power,
the world’s biggest democracy has entered a manifestly shaky period. But how
should we conceptually map the shakiness? My basic claim in this chapter is that
India after 2014 is not a case of democratic collapse but one of democratic ero-
sion or democratic backsliding.1 I will use the latter two terms interchangeably.
The analytical task of this chapter is twofold. What explains India’s democratic
longevity? And how might one explain the recent downward trajectory? My at-
tempt here is to provide an integrated argument, which seeks to answer both
questions. But before the argument is presented and to anchor the detailed dis-
cussion, it seems fitting to present a brief overview of India’s democratic record.
For most democratic theorists, competitive elections are a necessary condition
for the functioning of a democracy. “No elections, no democracy” is a theoretical
dictum of widespread acceptability. So let us begin with India’s electoral record.
Since independence in 1947, India has held 17 national and 389 state
elections. Power has changed hands eight times in the national capital and tens
of times at the state level. The latter phenomenon is by now so common that po-
litical scientists have stopped counting state-level government turnovers. Until
1992–93, the third tier of government—at the town and village level—was the
only unelected tier, but a constitutional amendment finally filled that gap, too.
Since the mid-1990s, roughly three million local legislators have been elected
every five years. Other than a twenty-one-month period of nationwide author-
itarianism (June 1975–March 1977) and a few electoral suspensions in areas of

India’s Democratic Longevity and Its Troubled Trajectory 35
unrest and insurgency, elections have decided who will rule India and its states
and, after 1992–93, its local governments as well. This has been true even in the
period of democratic backsliding since 2014. Several democratic institutions
have been challenged, causing the erosion, but the integrity of elections has not
been undermined.
Indeed, the idea that competitive elections are the only way to form
governments has been the institutionalized political commonsense of the
country. Such institutionalization means that for a long time now, no major
political actor or organization has proposed a non-electoral way of coming
to power. The question of whether there is more to democracy than elections
has remained unsettled and contested, but there is no doubt that competitive
elections have formed the core of India’s democratic imagination. It is hard to
predict whether the electoral core of democracy will remain unimpaired in the
coming years, but as of now, despite the ongoing democratic erosion, the elec-
toral principle remains intact. Modi may not have lost nationally since 2014, but
he has lost a number of state elections, which include states that are, politically
and economically, extremely significant. A Trump-like campaign, questioning
election integrity in the face of defeat, something not uncommon in many parts
of the world, has not been launched.
In the mid-1960s, Barrington Moore was among the first to note India’s dem-
ocratic credentials: “[A]‌s a political species, [India] does belong to the modern
world. At the time of Nehru’s death in 1964, political democracy had existed for
seventeen years. If imperfect, the democracy was no mere sham. . . . Political
democracy may seem strange in both an Asian setting and one without an in-
dustrial revolution” (Moore 1966, 314). Roughly half a decade later, in what has
become a foundational text of democratic theory, Robert Dahl identified India as
“a deviant case . . . indeed a polyarchy” (Dahl 1971, 68–69). About two decades
later Dahl was even more emphatic, calling India “a leading contemporary ex-
ception” to democratic theory (Dahl 1989, 253). Finally, after a little over another
decade, Adam Przeworski et al. (2000, 87) argued that in their 1950–90 interna-
tional dataset, India’s democratic longevity was the most surprising: “The odds
against democracy in India were extremely high.”2
A fairly substantial body of literature has sought to explain why India stayed
democratic for so long in a theoretically counterintuitive setting (Chhibber 2014;
Kohli 2001; Kothari 1970b; Moore 1966; Varshney 1998, 2013; Weiner 1989). In
this chapter, I engage the comparative or theoretical literature of a more recent
vintage, as well as probe the new datasets that measure democracy worldwide.
I advance two arguments. First, seeking a reexamination of how democratic
India has been, I draw a distinction between India as an electoral democracy and
India as a liberal democracy. Using political theory, India’s political history, and
the V-Dem dataset (Coppedge et al. 2021), I argue that India’s electoral record is
considerably better than its performance as a liberal democracy. India has, on the
whole, been electorally vibrant, but its democracy has substantial liberal deficits.
Under the twice-elected Modi regime (2014–present), these deficits have wid-
ened quite alarmingly. Substantially eroding civil freedoms, minority rights,
and institutional constraints on executive power, these deficits have primarily
affected the liberal side of democracy, not the electoral side.3 In my argument
below, I will call competitive elections a minimal democratic requirement, while
proposing that a fuller, or deeper, democracy also constrains governments be-
tween elections—by guaranteeing civil freedoms, protecting minority rights,
and viewing executive authority as institutionally checked and delimited. India’s
recent democratic erosion is about the latter, not about the former, meaning that
India remains electorally democratic but it has rolled back the democratic deep-
ening that was under way for decades.
Second, for explaining democratic longevity, my argument concentrates on
the primacy of elite choices, not on the structural or cultural determinants of
democratic longevity that several democratic theorists have privileged, though
not all. My focus on elites is divided into three parts: (1) the founding moment
and the formative period of democracy, (2) the period since the only nation-
wide collapse of democracy (1975–77) until 2014; and (3) the period of erosion,
though not collapse, since the rise of Modi (2014–present). In the first period,
I demonstrate how elite values played a big role in institutionalizing democracy.
In the second period, I argue that while values explain the behavior of a seg-
ment of elites, especially those who led some of the constitutionally given inde-
pendent institutions of oversight, such as the Supreme Court and the Election
Commission, a large section of political elites developed serious interests in the
persistence of democracy. The bedrock of values that launched the democratic
experiment acquired a serious core of interests. While dealing with the recent
democratic decline under Modi, I return to the explanatory salience of elite
values. Modi’s values, and those in power with him in Delhi, privilege Hindu
nationalism, or Hindu supremacy, over what India’s Constitution says, espe-
cially with respect to citizen freedoms, religious equality, and minority rights.
Right through, such elite choices, framed as values or interests, dominate my
explanation of democratic persistence, not structural or cultural determinants.
Though I find the formative period hugely significant and analyze it in con-
siderable detail, it is clearly not enough to leave the explanation at such strong
beginnings. Founding moments do not last forever; elite generations change
and their values evolve; and norms that got institutionalized can be broken,
as happened during India’s Emergency (1975–77) and might happen again.4
Origins and persistence are analytically distinct, and need to be separately
accounted for.
In the discussion below, using primarily the V-Dem dataset, I first examine
India’s democratic record, both over time and as compared to other countries.
Next, I analyze in detail the inadequacies of structural and cultural explanations
and, conversely, the superiority of elite-centric analytical accounts. I then turn
my gaze toward India’s democratic institutionalization in the early years of in-
dependence, dealing next with the return of a vigorous democracy (1977–2014)
after the Emergency breakdown, and turning finally to the erosion of democ-
racy with the rise of Modi after 2014. The concluding remarks recapitulate the
argument.
What the Statistics Say
The V-Dem dataset (Coppedge et al. 2021) has five democracy indices: electoral,
liberal, participatory, deliberative, and egalitarian. Of these, following standard
democratic theory, I stick to the first two: the electoral (for V-Dem, that means
free and fair elections, freedom of expression, and freedom of association5), and
the liberal (which covers, in addition to electoral democracy, individual and mi-
nority rights, and constraints on the executive, both legislative and judicial).
India’s Democracy over Time
Figure 2.1 presents India’s electoral and liberal indices since 1950. The electoral
index hovers around 0.7 for most of the period since then, with the exception of
the 1975–77 Emergency and the recent decline. The liberal index is consistently
below the electoral index, mostly staying between 0.5 and 0.6, with a lowering in
the two periods noted above: 1975–77 and in recent years.
India’s democratic record, thus, is marked by a consistent gap between the
electoral and liberal dimensions of democracy. It might be suggested that this is
not a specific ailment of Indian democracy. Rather, because of the way V-Dem
measures the two scores and the more stringent requirements of a liberal democ-
racy, the electoral–liberal hiatus is a general predicament of democracies. That is
indeed true, but the gap is more pronounced in India. For the period 1950–2019,
Figure 2.2 plots the decadal moving average of India’s electoral–liberal hiatus
against similar averages from some of the major world regions (Latin America
and the Caribbean, Asia-Pacific, and Western Europe and North America) as
well as the world at large. Compared to other regional and the world averages,
India’s gap is consistently larger, except for Latin America and the Caribbean
since the decade of 1990–99.6
1
0.9
0.8
0.7
0.6
Index
0.5
0.4
0.3
0.2
0.1
0
50
52
54
56
58
60
62
64
66
68
70
72
74
76
78
80
82
84
86
88
90
92
94
96
98
00
02
04
06
08
10
12
14
16
18
20
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
20
Year
Electoral Democracy Liberal Democracy
Figure 2.1. Electoral and liberal democracy index, India 1950–2020

Source: Coppedge et al. 2021
0.2
0.18
0.16
0.14
Electoral-Liberal Gap
0.12
0.1
0.08
0.06
0.04
0.02
0
1950−59 1960−69 1970−79 1980−89 1990−99 2000−09 2010−20
Decade
India Asia-Pacific Latin America and Caribbean Western Europe and North America World
Figure 2.2. The electoral-liberal democracy gap—decadal comparison

The legal foundations of India’s liberal deficits are explored later (in the sec-
tion entitled “New Elite Values and Democratic Decline”). Here it might be
pertinent to note some typical patterns that have repeatedly emerged in post-
British India.
Political parties are remarkably free at the time of elections. They are, of
course, prohibited from inciting violence. Short of that, they can make virtu-
ally any argument in election campaigns.7 Elections have been free and fair,
and their verdicts are respected by all including, crucially, the incumbents.
There have been arguments about how election finance might favor the
incumbents, but poorer parties have quite often won, especially at state level
but also at the national level.8 A party’s riches have not always helped, though
it is undeniable that incumbents typically have an upper hand in garnering
resources.
The election-time freedoms, however, coexist with their curtailments
between elections. Once the winning party or coalition forms the govern-
ment, restrictions are often placed on civil liberties, especially on freedom
of expression. Governments generally make two kinds of arguments against
free speech: that it offends the sentiments and honor of certain groups, or
it hurts national interest. On these two grounds, writers, artists, students,
and nongovernmental organizations have often been legally or adminis-
tratively harassed, even jailed. In a society marked by deep-rooted ascrip-
tive hierarchies, some group or another can always claim it has been hurt
by a speech, an article, a cartoon, a piece of art, a novel, a play, or a film.
When the argument about group injury or national interest enters politics,
governments rarely defend the writer, the intellectual, the filmmaker, the
artist, or the NGO.
These problems are common to all kinds of governments and parties.
Consider some of the biggest examples. In 1988, the Congress Party gov-
ernment banned Salman Rushdie’s Satanic Verses because of the protest of
Muslim right, which argued that sections of the novel, by insulting Prophet
Mohammed, had hurt the feelings of Muslims. Also under Congress Party rule,
the late M. F. Husain, a leading painter with a stellar reputation for artistic ex-
cellence, had to emigrate to the UAE for the last years of his life because the
Hindu right vociferously contended that his paintings of Hindu goddesses were
obscene, and the obscenity gravely offended the Hindu community. But there
is no doubt that the gulf between the electoral and liberal aspects of Indian de-
mocracy becomes especially wide when Hindu nationalists are in power. This
was true between 1998 and 2004 and has been especially glaring since 2014.
The final section (“New Elite Values and Democratic Decline”) takes up this
matter in detail.
More International Comparisons
Let us now turn to some more cross-regional and cross-country comparisons.

On the electoral democracy index, Figure 2.3 compares India’s performance with
the following averages: for the world, for Western Europe and North America,
for Asia-Pacific, and for Latin America and the Caribbean (Latin America here-
after). Generally speaking, after independence, India’s index has only been lower
than the average for Western Europe and North America, and higher than the
average for the world, Asia-Pacific, and Latin America, including the period
after the early 1980s, when democracy returned to Latin America. However,
there is one set of exceptions worth noting.9 After remaining ahead for all of
the post-1950 period, India’s index in more recent years has dipped below Latin
America’s—and has even fallen lower than the world average.
Figure 2.4 reproduces the preceding exercise on the liberal democracy index.
The similarities with the electoral democracy comparison are evident. India’s lib-
eral democracy index, with no exceptions, is lower than that for Western Europe
and North America. But, as above, it is on the whole higher than the indices for
Latin America and Asia-Pacific. And, as was true for electoral democracy, India’s
recent performance as a liberal democracy turns out to be worse than that of
Latin America. Indeed, its descent below the world average is also noticeable.
Figures 2.5 and 2.6 draw the comparison differently. Instead of comparing
India to regions, they place India’s indices against the well-known examples of
democratic longevity in the Global South. Costa Rica and Botswana are often
cited as the longest-surviving democracies in Latin America and sub-Saharan
Africa; and Przeworski et al. (2000) also list Jamaica, Mauritius, and Papua New
Guinea in the same category from their 1950–90 dataset.10 Figure 2.5 compares
all of these countries on the electoral democracy index: Costa Rica ranks con-
sistently higher than India since 1950 and Mauritius after its independence in the
late 1960s. Figure 2.6 compares their liberal democracy scores. Relative to India,
Costa Rica again scores higher since 1950, and Mauritius and Botswana since the
late 1960s. Also, India drops precipitously in more recent times.
In summary, the following inferences can be drawn from the empirical in-
vestigation above. First, India performs better as an electoral democracy than
as a liberal democracy. Second, until recently, compared to regional averages,
the electoral and liberal vigor of India’s democracy has been generally behind
only Western Europe and North America. Third, inter-country comparisons
show that Costa Rica has been consistently ahead since 1950, Mauritius does un-
exceptionably better since its independence, and democracy in Botswana, too,
displays greater liberalism in the last six decades. Finally, if Przeworski et al. are
right and India is still a very hard place for democracy—arguably more so than
1
0.9
0.8
0.7
Electoral Democracy Score
0.6
0.5
0.4
0.3
0.2
0.1
0
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Year
Asia-Pacific India Latin America and the Caribbean Western Europe and North America World
Figure 2.3. Electoral democracy index: India and some world regions, 1950–2020
1
0.9
0.8
0.7
Liberal Democracy Score
0.6
0.5
0.4
0.3
0.2
0.1
0
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Year
Asia-Pacific India Latin America and the Caribbean Western Europe and North America World
Figure 2.4. Liberal democracy index: India and some world regions, 1950–2020
1
0.9
0.8
Electoral Democracy Score
0.7
0.6
0.5
0.4
0.3
0.2
0.1
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Year
Botswana Costa Rica India Jamaica Mauritius Papua New Guinea
Figure 2.5. Electoral democracy index: India compared with selected countries, 1950–2020
1
0.9
0.8
0.7
Liberal Democracy Score
0.6
0.5
0.4
0.3
0.2
0.1
0
1950
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
Year
Botswana Costa Rica India Jamaica Mauritius Papua New Guinea
Figure 2.6. Liberal democracy index: India compared with selected countries, 1950–2020
Botswana, Costa Rica, Jamaica, Mauritius, and Papua New Guinea—it can only
be so because India’s income, given their income-intensive explanation of dem-
ocratic persistence, has been the lowest among all the longest-surviving dem-
ocracies of the Global South, which turns out to be true. In 2020, the per capita
incomes of these countries (in current US$) were: Botswana ($6,405), Costa Rica
($12,141), Jamaica ($4,664), Mauritius ($8,627), Papua New Guinea ($2,757),
and India ($1,928).11
Toward Explanation
Let us now examine India’s democratic persistence in light of the larger liter-
ature on democracy, theoretical and comparative. I divide up the explanations
into two blocs: structural and elite-centered. I first examine the applicability of
the mainstream democratic theory, which is mainly structural, to India. I then
move on to what appears to be more a promising explanatory route—the elite-
based explanations.
The Inadequacy of Structural (and Cultural) Explanations
A very large body of democratic theory has primarily pointed to structural factors
that make it easier, or harder, for countries to institutionalize democracy: level of
income (Lipset 1959; Przeworski et al. 2000), level of inequality (Acemoglu and
Robinson 2006; Boix 2003), class structure (Moore 1966; Rueschemeyer et al.
1992), and degree of ethnic diversity (Lijphart 1977; Mill [1864] 1975). Some
other scholars locate the explanation in the properties of political or societal cul-
ture (Almond and Verba 1963; Inglehart and Welzel 2007; Putnam 1993). If the
alternative line of inquiry is labeled elite-centered, then it is not altogether ana-
lytically illegitimate to stick both structural and cultural explanations together.
The latter explanations derive politics from the cultural structure, instead of the
economy, class, or ethnic structure.
There is no doubt that, generally speaking, the structural factors have some
validity. The correlation, for example, between high incomes and democratic
survival is considerable. But two sorts of reservations emerge right away. One is
simply epistemological. So long as we are not in the world of deterministic law-
like generalizations, but in a stochastic realm, which is where social science theo-
rizing normally belongs, the general claims about a group of states will not apply
to each individual state. For instance, the fact that low incomes are not normally
connected with democratic longevity does not mean that the relationship would
operate in each country. The statistical outliers will obviously not be on the trend
line. This distinction between what may be true at a group level, but not at an in-
dividual level, can easily be applied to India.
Przeworski et al. (2000) argue that income correctly predicted the type of re-
gime in 77.5 percent of the cases (Przeworski et al. 2000, 79). But it also means
that in 22.5 percent of the cases, the relationship did not hold. India clearly
belongs to this latter, smaller category. It had been a low-income country until
roughly 2005, graduating to a middle-income category after that,12 but a low per
capita income did not undermine its democracy.
Similarly, Acemoglu and Robinson (2006) may well be correct in stating that
“democracy has the best chance to emerge in societies with middle levels of ine-
quality” (37), and that neither in egalitarian nor low-inequality societies (such as
Singapore), nor for that matter in highly unequal societies (such as Sierra Leone),
can democracies stabilize themselves. But they quickly add that this prediction
“about inter-group inequality may not translate into statements about standard
measures of inequality and income distribution (e.g., the labor share or the Gini
coefficient)” (36–37). Given the statistical untranslatability of their concept of ine-
quality, it is hard to know whether India’s income Gini coefficient, estimated to be
between 0.35 and 0.45 since independence, is low, middling, or high. But even if it
is middling, in which case Acemoglu and Robinson could say India supports their
theory, it is not clear why India’s huge caste disparities should count for less than its
middling income inequalities. In Britain or Latin America, class realities might have
made universal franchise a means of obtaining the “right to a good coat . . . a good
hat . . . a good roof . . . [and] a good dinner” (24), but in India, securing dignity from
the daily insults of the caste system might well have been a larger impulse, as some
have argued (Mehta 2003; Varshney 2000; Weiner 2001). Consider how the lower-
caste Ezahavas were treated until the first three decades of the twentieth century.
They were not allowed to walk on public roads. . . . They were Hindus, but
they could not enter temples. While their pigs and cattles [sic] could fre-
quent the premises of temples, they were not allowed to go even there.
Ezhavas could not use public wells or public places. . . .
An Ezhava should keep himself, at least thirty- six feet away from a
Namboodiri Brahmin and twelve feet away from a Nair. . . . He must address
a caste Hindu man, as Thampuran (My Lord) and woman as Thampurati (My
Lady).. . . He must stand before a caste Hindu in awe and reverence, assuming
a humble posture. He should never dress himself up like a caste Hindu; never
construct a house on the upper class model. . . . [T]‌he woemen [sic] folk of the
community . . . were required, young and old, to appear before caste Hindus,
always “topless.” (Rajendran 1974, 23–24)
In conditions like this, lasting for centuries, the struggle of lower castes in
India has been, first, for dignity and for being treated as human beings, and only
later for higher income equality (Chakrabarti 2019). Indeed, lower castes, given
their large numbers, have used India’s democracy to press somewhat successfully
for conditions of dignity; their success at income improvements have not been as
great (Chauchard 2017; Jensenius 2017; Varshney 2000).
Let us now turn to the arguments based on class structure. If Barrington
Moore (1966) is right about the “no bourgeois, no democracy” theoretical prin-
ciple, or if it is generally true that the rise of middle classes moderates distribu-
tional struggles between the rich and the poor (Acemoglu and Robinson 2006),
thereby making greater structural room for democracy than would be true oth-
erwise, then it is noteworthy that India’s middle class was minuscule at the time
of independence, perhaps no more than 4–5 percent of the population, and it is
only after the great economic turnaround over the last three decades that India
has developed a substantial middle class, now widely believed to be about a third
of the country. Yet democracy kept functioning even before the 1990s economic
reforms increased the size of the middle class.
Political culture and ethnic diversity as explanatory factors run into sim-
ilar difficulties. The political culture argument, both in its earlier (Almond
and Verba 1963) and later forms (Inglehart and Welzel 2007; Putnam 1993),
concentrates on the fit between certain enduring social or cultural character-
istics and the type of polity. On the whole, it is argued that a more egalitarian
social structure generates civic culture, or horizontal social capital, which
makes democracy work. India’s caste system, the defining feature of Indian so-
cial structure for centuries, is the antithesis of equality. It has viewed human
beings, and the caste groups to which they belong, as fundamentally unequal.
Along with racial stratification, the hierarchical caste system has been termed
a prototypical form of vertical social order (Horowitz 1985; Varshney 2012).
Such verticality has generally ruled out the idea and experience of human
equality. In rural India, the caste system is more entrenched than in the cities,
but even as late as 2011, India was 68 percent rural and as much as 83 per-
cent in 1951. The hierarchies of the caste system have not prevented democracy
from working in India.13
The argument that ethnic diversity is an unfavorable condition for de-
mocracy goes back to John Stuart Mill. Mill’s original argument took the fol-
lowing form:
Free institutions are next to impossible in a country made up of different nation-

alities. Among a people without fellow-feeling, especially if they read and speak
different languages, the united public opinion, necessary to the working of rep-
resentative government, cannot exist. (Mill [1864] 1975, 384; emphasis mine)
Thus, multiple languages, according to Mill, would make the creation of a

nation highly improbable. A superordinate loyalty to a national center is nec-
essary, or free institutions would simply fracture societies and thereby under-
mine democracy itself. Arend Lijphart (1977) developed a creative solution to
the problem Mill posed by arguing that democracy in multiethnic societies was
actually possible if it took a consociational, not a liberal, form.
India has undoubtedly departed from Mill’s expectations. The country has fif-
teen languages spoken by more than ten million each, yet democracy was not
stymied by a multi-linguistic society. Interregional communication across the
nation has been possible because India developed a three-language formula for
its school education, allowing citizens in different parts of the country to under-
stand each other. As a result of the education policy, India’s literates became bi-
lingual or trilingual.14 Moreover, multiple languages notwithstanding, India has
managed to generate considerable loyalty to the national center, something Mill
considered virtually impossible (Stepan, Linz, and Yadav 2011).
If Mill’s apprehension has not come true, how does India measure up to
Lijphart’s reasoning about ethnic diversities and democracy? Two responses
can be given. The first is strictly formal. In a formal or literal sense, India defies
Lijphart’s consociational model. Its liberal deficits notwithstanding, India has
mostly been a liberal democracy, not a consociational one.15 Individuals, rather
than groups, have been the fundamental unit of political representation. Indeed,
during the break-up of British India in 1947, a serious consociational argument
was made in mainstream politics. In a consociational style, the Muslim League,
which led the Pakistan movement, demanded that it be recognized as the only
political party representing the Muslims of India, and that there also be a fixed
share of legislative, administrative, and judicial seats reserved for Muslims.16
The Congress Party, leading the independence movement, chose to take the risk
of partition rather than accept the Muslim League’s consociational demands to
alter the fundamental nature of the polity. It went for a type of democratic polity,
in which the individual would be the fundamental unit of political arrangements
(Khosla 2020), with no political organization acting as a “sole spokesman” for
any community, and no fixed religious quotas in the institutions of governance.17
It argued in favor of an open interparty competition for each community’s votes.
And there would be no grand coalitions or minority vetoes. Minority rights—for
education and religious personal laws for marriage, divorce, property inherit-
ance, and so forth—would be constitutionally granted, but no minority quotas in
government, parliament, or civil service would be permitted.
The second answer, however, is more about political practices that depart
from the liberal model of individual-based representation and give India some
power-sharing features. The two most important are federalism and caste-based
reservations. India’s federalism is linguistic, in that all major linguistic groups
have a state of their own in the federation. And the lowest castes in the Hindu
hierarchy have guaranteed representation in legislatures as well as bureaucracies,
in proportion to their weight in the population.
But neither of these features makes India a power-sharing democracy in a
consociational sense. First, federalism is linguistically based, not religious. For
over a century, if not more, religion has been the greatest fault line of Indian pol-
itics. Consociational polities are typically formed around the deepest cleavage
patterns, but independent India rejected power-sharing on the basis of reli-
gion. Instead, it conceded power-sharing on linguistic lines. Second, the caste-
based reservations were primarily for legislatures and civil services, not for the
executives and cabinets, which is where most power resides.18
Bringing Political Elites In
If structural accounts are not fully adequate, where do we go next? Agent-centric

explanations have been the standard alternative path to follow. In the main,
the agents that carry democratic politics include elites and political organiza-
tions, especially political parties, and those related to movements. These are also
often—and rightly—labeled as political, as opposed to structural, factors.
Some of this analytic spirit is evident in two relatively recent works on democ-
racy. Mainwaring and Pérez-Liñán (2013) write:
Political actors, not structures or cultures, determine outcomes, even though

structures and cultures affect the formation and preferences of actors. We view
presidents and organizations such as parties, unions, business associations, the
military, and organized movements as the most important actors. . . .
We emphasize the role of political factors that help political regimes survive or
lead them to fail. By “political factors” we refer specifically to the impact actors’
normative preferences about democracy and dictatorship . . . exercised. . . . We
counterpose an emphasis on these political factors to analyses that argue that
the survival or displacement of regimes depends largely on structural factors
such as the level of development, the class structure, or income inequalities, or
on mass political culture. (Mainwaring and Pérez-Liñán 2013, 5)
Ziblatt (2017) also points to the importance of elites. After noting that
the three most widely cited accounts of democratizations have concen-
trated on (1) the level of economic development, (2) the emergence of a
middle class, and (3) the demands of the working class for greater rights
and a larger share of power, he favors the elite-c entric account of O’Donnell
and Schmitter (1986):
If elites can be made secure with regard to their future wealth, status, and power
as democratization unfolds, . . . critical portions of the elite will become reluc-
tant but essential democrats. On the other hand, if elites remain insecure, polit-
ically fragmented, and fearful of major loses [sic] in their future, then they will
support and even lead counterrevolution, creating a historic record of unsettled
democratization. (Ziblatt 2017, 16)
Stating further—and rightly—that “a country’s political regime is not simply

a mirror of its economy” (17), Ziblatt concentrates on political parties as actors,
arguing that “it is crucial to elevate political parties, the ‘political carriers’ of or-
ganized interests, to the status of a variable that shapes democratization” (20).
This line of reasoning invites a serious examination of what the political elites
want and why. But we still have an important larger issue left to consider: should
we focus on elite values or elite interests? It is not clear if we can answer this ques-
tion generally, theoretically, or in the abstract. It may be an empirical, not a the-
oretical, matter.
Ziblatt’s elite-centered argument is strictly about interests. European materials
show, he argues, that if the incumbent elites, customarily represented by conser-
vative parties, came to the view that democratization would hurt their power or
wealth, they impeded its evolution, and whether or not that brought democrati-
zation to a halt at its moment of inception, democracy over the medium to long
run stumbled as, for example, in pre-1945 Germany. A cooptation of conserva-
tive parties—and thereby the interests of traditional elites—augured best for a
smooth democratic trajectory, as in post-1832 Britain.
Central to this argument is the idea of a clash between incumbent elites and
emerging elites, each represented by a different political party, and a compro-
mise between the two as the foundation of democratic stability. Whatever the
relevance of this argument for Western Europe in the nineteenth century, its ap-
plicability to post-colonial democracies of the twentieth century is not straight-
forward. Colonialism accounts for the greatest political difference between
nineteenth-century Europe and the European colonies in the tropics. Generally
absent in the colonizing Europe, anticolonialism was a big political sentiment
in the colonies. The latter gave political parties and groups fighting for indepen-
dence extraordinary legitimacy. Local elites allied with the colonial masters,
however socially powerful, were eventually no match for the anticolonial organ-
izations and parties.
Furthermore, the colonial powers, even those that were democratic like
Britain, argued that the colonies were unfit for democracy (see Mehta 1999; Mill
[1864] 1975). As a result, in a late colonial setting, the biggest adversary of democ-
racy was often the colonial power, not the conservative landlord pitted against a
democracy-carrying middle class or the bourgeoisie, as in nineteenth-century
Europe. In countries like Pakistan, which came about as a result of a separatist,

not an anticolonial, movement, feudal landlords could still be viewed as the in-
cumbent elites, who had to assess their interests in an incipient democracy, as
Tudor (2013) has argued. But where colonial departure was not accompanied by
the colony’s partition, the colonial rulers were, more often than not, the greatest
political force opposed to democracy.19
Let us nonetheless suppose that the indigenous princes and landlords, key
allies of the British, can be viewed as incumbent elites in India by virtue of their
association with the colonial rulers. In Britain, as Ziblatt shows, the Conservative
Party represented their interests and it had the power to block democratization,
unless concessions were made by emerging elites. In India, conservative parties
did emerge to protect the landlords in the late colonial period, but generally
speaking the emerging elites squarely defeated them in the provincial elections.20
Those who did not participate in the independence movement simply had no
effective political space to be counted as powerful political actors. The Congress
Party attacked landlordism as early as 1936–37, renewing the assault after inde-
pendence, when it was firmly in control of state power.21
But there is a sense in which the claim that, for democracy to evolve smoothly,
it is necessary to accommodate the interests of incumbent elites can be made for
India as well. First, after independence, when the Congress Party enacted a land
reform program, aimed at crushing landlord power, the landlords began to op-
erate at a regional, vernacular level. As Weiner (1967) showed, the landlords pen-
etrated the district and provincial levels of the Congress Party as party members,
and managed to block land reforms with varying degrees of success in different
states. In an India which was 83 percent rural at the time of independence, the
Congress Party needed the landlords to expand its organizational presence in
the countryside, as it is the landlords who were locally influential, not the peas-
antry, which was dependent on them for a whole variety of economic and social
reasons and, therefore, quite powerless (Herring 1982). Basically, there emerged
a contradiction between the political imperatives of party building and the eco-
nomic policy goals of land redistribution and tenancy reforms (Varshney 1995).
And once the landlords were inside the party, they could begin to impede the im-
plementation of land reforms. Thus, conservative parties representing landlord
interests might not have risen in a big way, as in Western Europe, but a function-
ally equivalent political form—penetration of the ruling party at lower levels—
emerged to protect landlord interests.22
But these messy equations emerged later, not at the time of India’s demo-
cratic commencement. Moreover, they affected the lower political levels. The de-
cision to democratize was taken at the summit of the polity, which was firmly
in the hands of the emerging, not incumbent, elites.23 It is the emerging elites
who devised India’s democratic constitution. And in the vigorous debates of the
Constituent Assembly (CA), lasting nearly three years, one sees enough evidence
of arguments based on values, not any recognizable display of interests.
One could, of course, say that the distinction between elite values and elite
interests was only theoretical at the time of independence, given the Congress
Party’s hegemony. It could afford to talk in terms of values, for the party could
not possibly visualize losing power. Establishing a democracy, therefore, could
not conceivably hurt.
To be sure, it was not easy for the Congress Party to envisage a defeat at the
hands of the opposition in the very near future, but a constitution is a statement
about the enduring properties of a polity, not a device for short-term power dis-
tribution. Accepting the idea of elections meant that the party could lose them
in the future. Indeed, powerful opposition parties had already appeared in South
India. And within ten years of independence, the Congress Party would lose
the state of Kerala to the opposition parties there (and another ten years later, it
would be defeated in many other states).
Interests attached to democracy did emerge later at higher levels of polity. But
at its inception, the tenor of debates about democracy do not provide evidence
of interests. That is why it is important to draw a distinction between the origins
of democracy in India, and its persistence. The former was heavily based in elite
values, and the latter in a combination of values as well as interests, as we shall
note later (see the section, “Democracy’s Second Innings and the Birth of Elite
Interests”).
The Emergence of Democracy and Elite Values
India’s post-independence elites began their political education in 1920, not

at the time of independence in 1947. The former was the so-called Gandhian
turning point in India’s freedom movement, when the Congress Party, which
led the movement, turned toward mass politics. Under Gandhi’s leadership, it
stopped being a lawyer’s club, making legal arguments with the British in the
Queen’s English. Instead, it started mobilizing the vernacular masses against co-
lonial power. The mass-based movement lasted nearly three decades before in-
dependence came. What were the values the movement promoted and the elites
imbibed? How did those values shape the creation of democracy?
Let me first examine the elite views about elections and universal franchise.
I will provide evidence with respect to (1) the entire class of political elites,
where it is available. Where it is not, I will concentrate on (2) the ruling Congress
Party, which lost elections in only one state in the first two decades of indepen-
dence (but started losing in several states thereafter), and (3) Jawaharlal Nehru,
India’s Prime Minister from 1947 through his death in 1964, the winner of three
consecutive national elections—in 1952, 1957, and 1962—and the preeminent

political figure of the first seventeen years of Indian independence. He did not
always win policy or institutional struggles, but his stamp on the evolution of
the democratic structure is beyond doubt. I will, therefore, combine three kinds
of arguments: those at the level of all political elites in general, those dealing
with the ruling Congress Party, and those covering the very summit of political
leadership.
Universal Franchise
After nearly two hundred years of British rule, India’s literacy rate at indepen-
dence was a mere 17 percent. More than half of the country was below the pov-
erty line (Ahmed and Varshney 2012).24 The vastly poor and illiterate masses
were in no position to demand democracy. Yet Indian elites committed them-
selves to universal franchise. Indeed, the commitment was made by the indepen-
dence movement as early as 1928, the same year as universal franchise came to
Britain.25
In the last decades of British rule, starting in 1921, Indians were allowed
to vote, but franchise had three limiting aspects. First, voting was not a right
but a privilege available on the basis of income, literacy, and landownership.
Second, the first arena of voting was local government and, beginning in 1937,
it was also extended to provincial assemblies. Before the British left in 1947,
India had had two provincial elections, in 1937 and 1946. But at no point until
their departure was the central legislature or government elected. Moreover,
the highest British officials of the land could veto the legislation of provin-
cial assemblies and executive acts of provincial governments. Finally, the elec-
torate numbered thirty million in 1946, roughly 15 percent of the total adult
population, but the entire country did not constitute a voting college. Rather,
somewhat like what came to be called consociational democracy later, the
British viewed India as a polity based on communities, not individuals. In
particular, the electorate was divided on the basis of religion, with separate
electorates for minorities. In such “separate constituencies,” only the minor-
ities could vote and only members of minority communities could run for
office. Functional representation was also added to this legislative scheme.
Some legislative seats were reserved for trade unions, business groups, the
princely states, and so on.26
The freedom movement sought a radical overhauling of this institutional
framework. One has to accept the great twentieth-century premise, argued
Nehru, that “each person should be treated as having an equal political and social
value” (Nehru 1942, 67). This became the defining statement about independent
India’s universal franchise. It did not promise economic equality, only political
equality. But that was enough to make the case that gender, income, property
ownership, literacy, caste, religion, language, and tribe could not be the basis for
allocation of voting rights. For decisions on who would rule India, voting was
essential and all citizens had the right to vote. Nehru never gave up his belief in
elections as the only way to determine which party will govern every five years.
“Elections were an essential and inseparable part of the democratic process and
there was no . . . doing away with them” (Nehru 1946, 53). He would also argue
later: “Our democracy is a tender plant which has to be nourished with wisdom
and care and which requires a great deal of understanding of its real processes
and its discipline.”27
Nehru, of course, was not alone. In 1945, an important committee of the
Congress Party argued against those who thought that the poor and unedu-
cated voters would make too many mistakes and would not be able to use the
right to vote wisely and well. Although the voter’s “judgement may be faulty,
his reasoning inaccurate and his support of a candidate not infrequently de-
termined by considerations removed from a high sense of democracy, he is yet
no better or worse than the average voter in many parts of Europe where adult
franchise has been in force for some time” (Sapru Committee Report, cited in
Austin 1966, 147).
Nor was the sentiment confined to the Congress Party. In the early years of
independence, the most important institutional site for India’s political elites was
the Constituent Assembly, which worked for over three years, from late 1946 to
late 1949, to draft India’s Constitution, a document that has survived until today.
Consisting of 299 members, the Constituent Assembly was elected by provin-
cial assemblies, and though most members were from the Congress Party, it also
had several members whose ideology was different and who later became great
critics of the ruling party. The latter group included some Hindu nationalists,28
as well as the famous Dalit leader, B. R. Ambedkar, who was made the head of
the Constitution drafting committee.29 A Dalit leader, a bitter critic of Gandhi,
thus became a father of India’s Constitution. Ambedkar made a famous argu-
ment about why even literacy was not a condition for franchise which, he argued,
had to be universal.
Those who insist on literacy as a test and insist upon making it a condition
precedent to enfranchisement, in my opinion, commit two mistakes. Their
first mistake consists in their belief that an illiterate person is necessarily an
unintelligent person. . . . Their second mistake lies in supposing that literacy
necessarily imports a higher level of intelligence or knowledge than what the
illiterate possesses. (Ambedkar’s presentation to the Simon Commission, 1928,
reproduced in Jaffrelot and Kumar 2018, 34)30
The claim here was not that the illiterate people should not be educated; only
that illiteracy and intelligence are analytically separable, and even the illiterate
understood their interests. The chair of the Constitution Drafting Committee
and his firm voice mattered.
There were undoubtedly some voices of dissent. Some members continued
to favor voting restrictions, especially those based on literacy. A member
pleaded: “For the first ten years, just limit this right of voting to literate
people .. . . Otherwise, in my humble opinion, these elections will be a great
farce. . . . My submission is that . . . we should have the provision of literacy”
(Lok Sabha Secretariat 1949a).
But such views were few and far between. They could not win the day. With an
overwhelming majority, across the political spectrum, the Assembly embraced
universal adult franchise. The reasons for support were not identical,31 but the
support was nearly unanimous. And the final conclusion was unmistakably
clear. “The Assembly . . . adopted the principle of adult franchise . . . with an
abundant faith in the common man . . . and in the full belief that the introduc-
tion of democratic government on the basis of adult suffrage will . . . promote the
well-being . . . of the common man” (Lok Sabha Secretariat 1949b).
As a consequence, India’s electorate expanded from 30 million in 1946 to
173 million in 1951–52. The first elections, based on universal franchise, took
four months, starting in October 1951 and concluding in February 1952. They
were the biggest election exercise in history.32 Two more elections were held
before Nehru’s death (1964)—in 1957 and 1962. The idea of elections as the
only way to come to power deepened, becoming political common sense. The
Congress Party could have used Nehru’s death (and later a war with Pakistan in
1965) to suspend the idea of elections and announce that it was in power for the
foreseeable future, if not in perpetuity. No arguments of this kind appeared in
the political sphere. After Nehru’s death, the fourth general elections, covering
national parliament and all state assemblies, took place in 1967. The Congress
Party did not lose power at the national level, but it was defeated in several states.
It bowed out wherever it was defeated. It is only in 1975 that the deepening insti-
tutionalization was ruptured. This is discussed below.
Democracy’s Second Innings and the Birth of Elite Interests
In June 1975, Indira Gandhi, prime minister since 1967, declared a state of emer-
gency and suspended democracy. It was a brief interlude, lasting twenty-one
months.
Two analytical questions are relevant concerning the 1975–77 rupture. Why
did a break, however short, come about? And what was done to keep democracy
alive after the break? The first question is about the conditions under which dem-
ocratic institutionalization crumbled, the second about its return. Paradoxically,
in both, the Constitution played an important role. But it was not the only factor.
The Constitution, an expression of elite values to begin with, started engaging
elite interests in the mid-1970s and after.
Faced with a movement against her government, which had been under way
for some time and was invigorated further when a High Court nullified her elec-
tion victory on the grounds that her campaign expenses exceeded the ceiling
prescribed by law, Indira Gandhi threw out democracy altogether in June 1975.
Her regime arrested over 110,000 opposition leaders and activists, inaugurated
the doctrine of the executive supremacy in lawmaking, curbed the power of judi-
cial oversight, imposed press censorship, abrogated citizens’ right to free speech
and assembly, and allowed detention without trial.33 All of this was done using
Article 352 of India’s Constitution, which allowed suspension of routine dem-
ocratic processes if “a grave emergency exists whereby the security of India is
threatened by internal disturbances.” In other words, the Emergency was anti-
democratic, but not unconstitutional. A Supreme Court judgment later, for all
practical purposes, certified its constitutionality, when the Court said that deten-
tion without trial and a lack of judicial scrutiny of the executive and legislature
were justified in a state of emergency.34
Eighteen months later, in January 1977, Indira Gandhi announced new
elections, releasing jailed leaders and activists and restoring citizen and press
freedoms. She was seeking to re-legitimize her rule. The best available hypothesis
for why she did so suggests that even she found it hard to devise non-electoral
arguments for continuance in power (Weiner 1989). She lost the elections, which
were supposed to bolster political legitimacy, and bowed out after defeat, instead
of canceling election results as an all-powerful head of government, something
that has happened in many countries.35 In March 1977, the first non–Congress
Party government was formed in Delhi. (Since then, seven more non–Congress
Party coalitions have run the national government.)
As a new government took charge, one of its most important activities was to
ensure that the Constitution was not used to proclaim an internal Emergency
again. This was no longer a question of values alone. Given that virtually all
new cabinet ministers and their umpteen colleagues were jailed during the
Emergency, it was in the interests of the newly empowered elites to amend the
Constitution in such a way that a suspension of democracy in the future would
become inordinately difficult and they would not be jailed for political reasons.
The 43rd and 44th Constitutional Amendments sought to achieve this purpose.
“Armed rebellion” replaced “internal disturbance” as the basis for declaring an
Emergency which, from then on, would also require not just the recommenda-
tion of the prime minister to the president, the titular head in a parliamentary
system, but also the written advice of the cabinet, and would have to be endorsed
by a two-thirds majority in parliament within a month of the proclamation.
Moreover, the extension of an Emergency beyond six months would also require
parliamentary approval, not simply executive wish. Finally, the power of the ju-
diciary to investigate and judge the constitutionality of parliamentary legislation
and executive decrees was restored.
If it was in the interest of Indira Gandhi and her colleagues in 1975 to use
the Constitution to disable democracy, it was now in the interest of the new
rulers and their parties to make democracy’s suspension awfully difficult. “No
Emergency ever again” was the political motto. The constitutional amendments,
aimed at averting that eventuality, have stayed until now.
Indeed, India’s Constitution has not been overturned at all, only amended
from time to time. The Constitution has become an institutional bedrock
of India’s democracy. For India’s democracy to end, the Constitution would
have to be terminated, or radically amended to restore at least two ideas of the
Emergency: executive supremacy unrestrained by judicial scrutiny, and deten-
tion without trial.
India’s overall constitutional stability leads to two questions. Why has
the Constitution not been overturned, only amended? And through which
institutions does the Constitution maintain its centrality? Just because
the Constitution exists does not mean that it can exert political power.
Constitutions have repeatedly broken down, or been overthrown, in the de-
veloping world, which includes India’s neighborhood (Pakistan, Bangladesh,
Sri Lanka).36
Here, the fact that India consistently pursued a parliamentary system
appears to have come to its democratic rescue. Accumulating comparative ev-
idence makes this point clearer than ever before. In a widely read study, Stepan
and Skach (1993) surveyed all democracies between 1979 and 1989. The world
had forty-three consolidated democracies in this period, of which as many as
thirty-four were parliamentary and only five were presidential. After analysis,
they concluded that there was a “much stronger correlation between demo-
cratic consolidation and . . . parliamentarism than democratic consolidation
and . . . presidentialism” (Stepan and Skach 1993, 5).
A more recent study, by Przeworski (2019), comes to a similar conclusion,
though it approaches the question of longevity via its opposite: brittleness.
Between 1918 and 2008, of the forty-four consolidated parliamentary democra-
cies, only six fell, which gives us a mortality rate of 1 in 7.3. In the same period,
there were twenty-six consolidated presidential democracies, of which seven
fell, which yields a mortality rate of 1 in 3.7. That is twice as high as for consoli-
dated parliamentary systems. “The weakness of presidential systems,” concludes
Przeworski, “is manifest” (Przeworski 2019, 35).37
Presidential democracies have both repeatedly broken down and often sus-
pended constitutions. Corrales (2018) notes that since the early 1980s alone,
when democracy returned to Latin America, there have been twenty-four
attempts to elect Constituent Assemblies, all aimed at rewriting the entire consti-
tution; eleven were successful, thirteen were aborted.
There is a good deal of literature on why parliamentary democracies and their
constitutions are more durable.38 The details of the literature need not detain us
here, except to note a major recurring theme especially relevant to India. Being a
parliamentary system, which organically connects the executive and the legisla-
ture, India has never witnessed politically crippling stalemates between the two
separately elected summits of institutional power in a presidential system: the
president and congress. Such institutional paralyses, it is argued, have led to
more frequent democratic breakdowns in presidential systems. India’s adoption
of a parliamentary system and its ability to resist the temptation for a presidential
conversion of its polity appear to have a lot to do with its constitutional as well as
democratic longevity.
Of course, in and of themselves, constitutions do not act. To make their pres-
ence felt in active politics, constitutions require actors and institutions to stand
up and extend fighting support. It is necessary to understand who these agents of
constitutional and democratic stability were, and why.
Three kinds of actors have played this role in Indian democracy, espe-
cially after its awkward turn in 1975–77. First, armed with two constitutional
amendments described above (43rd and 44th), which restored its power of ju-
dicial review, the judiciary acquired new teeth in the post-Emergency era. If the
courts in the past assertively exercised judicial scrutiny of the executive and the
legislature, those elected to power would often choose the method of constitu-
tional amendment, or parliamentary supermajorities, to overturn unpalatable
judgments. However, after the executive excesses of the Emergency and their
electoral rebuff, the political space for an executive pushback declined for many
years thereafter and the elected elites did not think it advisable to push back
against the judges.39
In particular, the judiciary added two new forms of scrutiny for checking
executive and legislative power. First, Public Interest Litigation (PIL) was in-
stitutionally promoted. In conventional citizen–government interactions in a
democracy, the standard notion of locus standi allows only those aggrieved—by
executive or legislative action—to challenge the injury in a court. This require-
ment often works to the disadvantage of the poor and the marginalized because
they either might not have the resources to reach courts directly, or might not
even know that courts could invalidate government action. By allowing individ-
uals or organizations to argue on behalf of those who can’t afford litigation or are
unaware of its possibilities, PIL made a huge intervention into the democratic
political process, repeatedly pushing governments on behalf of the citizenry.

This practice took off in the early 1980s.40
Second, in arenas of national life, where no prior laws existed, the judiciary
adopted a quasi-lawmaking role, or initiated court-supervised executive action.
Using this new mode of intervention, the judiciary issued to the governments
guidelines on how to deal with sexual harassment in the workplace, on the rights
of sexual minorities and slum dwellers, and on a whole host of environmental
matters, including air and river pollution, forests, and wildlife.
Polls repeatedly showed in the 1990s and 2000s that the courts were more
popular than parliaments, assemblies, and political parties (Krishnaswamy and
Swaminathan 2019). This situation may not last forever, as the relationship be-
tween the judiciary on one hand and the executive/legislature on the other tends
to be more pendulum-like than static in democratic polities. But it is important
to note that the popularity of courts lasted long enough in India for the judi-
ciary to check executives and legislatures on behalf of the citizenry, or what has
sometimes been presented as national interest. While the independence of the
judiciary might have come under a cloud since the rise of Modi, as analyzed in
the next section, its power in the three-and-a-half decades after the Emergency
unquestionably increased. The term “judicial sovereignty” was used to describe
the power of courts after the 1990s (Mehta 2007).
The Election Commission is the second institution which started taking
its power seriously. Unlike the US and perhaps several other polities, India’s
Election Commission is set up by the Constitution as an independent institu-
tion. The Election Commissioners are appointed by the executive for a fixed
term, which can’t be altered by governmental changes. The primary responsi-
bility of the Election Commission is to conduct free and fair elections, which
includes preparation of election rolls, registration of political parties and elec-
tion contestants, watching election campaigns, checking voter intimidation,
and supervising the behavior of election officials. As McMillan (2010) has
argued, the Election Commission has gone through three phases in its evolu-
tion: a phase of confident establishment during the first elections under Nehru
(1947–64), a phase of subservient quiescence under Indira Gandhi (1967–84),
and a phase of assertive activism after the late 1980s. Modi’s rise may have
raised some doubts about the continuing independence of this institution, too,
but there is no denying the fact that it has maintained the integrity of the elec-
tion process for very long.
The quiescence of the middle period, listed above, was because Indira Gandhi
exercised fearsome power, and the constitutionally independent institutions
were either unwilling or unable to stand up to her. This included both the ju-
diciary and the Election Commission. And the activism of the last period was
driven by two interconnected factors, both leading to the greater consolidation
of the idea that free and fair elections were the only device for coming to power
and forming governments.
The first factor was the end of Congress Party hegemony and the emergence
of a fluid party system, leading to coalition governments, which lacked the over-
whelming power of an Indira Gandhi. And the second factor was the continued
popularity of the Election Commission, repeatedly evident in polls. Generally
speaking, the Election Commission and the judiciary kept receiving the highest
trust rating in polls during the 1990s and 2000s (de Souza, Palshikar, and Yadav
2008). As a result, much like the judiciary, the Election Commission also felt
bold enough to discipline both election officials whose behavior it deemed un-
fair and prejudiced and politicians whose campaign violated the legal canons, or
widely accepted norms. When it suspected voting fraud, it also “countermanded”
elections in some constituencies, ordering re-polls. Like the courts, the assertive
behavior of the Election Commission may also not last forever, but the asser-
tion has lasted long enough to generate strong citizen faith in the freedom and
fairness of elections. For decades now, no significant challenges to the election
verdicts have been launched, and the incumbents who lose elections routinely
bow out.
A third set of actors, in addition to the courts and the Election Commission,
is simply the political parties. So many parties have been in power since the early
1980s, mostly at the state level but also in Delhi, that serious stakes in the persist-
ence of democracy have been created. Parties not in power in Delhi have some-
times run half of the state governments. As a result, if the ruling party at the
center—and sometimes at the state level, too—egregiously breaks democratic
norms, countervailing power is, in principle, available in the system to oppose
the violation and seek to correct it, substantially if not wholly. Parties mobilize
citizens for mass protests, launch court challenges, and mount press campaigns.
This does not always, or fully, prevent the abuse of political power or eliminate
predatory governmental conduct, but the generally available balance of power
means that either the excesses can be reduced in frequency or their severity can
be checked. All political parties swear by the Constitution, and use that as a re-
source to challenge governmental excesses.
More important for our purposes, democracy could not easily be suspended,
for it would constitute the kind of executive excess that would touch off legal
challenges by political parties and countermobilization sponsored by them.
Democracy’s suspension would hurt the interests of simply too many elites, and
they have so far had power in the system to hit back.
Another way of conceptualizing this phenomenon is to say is that by giving
a taste of power to a large number of political parties, federalism in effect has
become a mainstay of democracy. Democracy would potentially be under a se-
rious threat if the same party ruled in Delhi and all states, unless the party was
committed to democratic values, not to perpetuation in power by hook or by

crook. A multiplicity of parties in governments, at the center and states, creates
countervailing power in the polity, if not an exact balance of power. As argued in
the following section, federalism has become a real problem for Modi’s desire for
greater control of the polity; for a collapse of India’s democracy, if not its erosion,
federalism would also have to come apart.
New Elite Values and Democratic Decline:

Modi and Hindu Nationalism
In its 2021 report, Freedom House announced: “India, the world’s most pop-
ulous democracy, dropped from Free to partly Free status” (2021, 2). And
putting its finger on the heart of the matter, it attributed the decline to the
Hindu nationalist ideology of Narendra Modi, India’s prime minister since
May 2014, who has won two straight national elections, in 2014 and 2019.
“Under Modi, India appears to have abandoned its potential to serve as a
global democratic leader, elevating narrow Hindu nationalist interests at the
expense of its founding values of inclusion and equal rights for all” (Freedom
House 2021b, 8).
V-Dem Institute’s 2021 report went further. Instead of calling India “partly
free,” it concluded that India in 2020 had become an “electoral autocracy” (V-
Dem Institute 2021, 6). Its judgment is based on two criteria. First, “the overall
freedom and fairness of elections . . . was hard hit” (20). And second, the
“diminishing of freedom of expression, the media, and civil society have gone the
furthest. The Indian government rarely, if ever, used to exercise censorship . . . be-
fore Modi became Prime Minister” (20). Hungary and Turkey, it adds, “became
(electoral) autocracies in 2018 and 2014 respectively, and India now joins their
ranks” (20).
V-Dem Institute is right about the second factor—the attack on freedom of
expression, media, and civil society—but wrong about the first, namely, the cor-
rosion of free and fair elections. Even under Modi, elections continue to be com-
petitive. Ironically, a remarkable piece of evidence came just a few weeks after the
publication of the V-Dem report in March 2021.
In West Bengal, Tamil Nadu, Kerala, and Assam, state elections took place.
The BJP in general and Modi in particular fought very hard. They lost all states
except Assam, but did not challenge the results. West Bengal was especially
important. It is the third largest state of India; its historical significance is be-
yond doubt; and of all states, Modi campaigned most vigorously there, leaving
Delhi twenty times in the space of roughly six weeks to address election rallies.
When the results came, a regional party under a political archrival had clearly
defeated the BJP. Instead of questioning the integrity of election results, Modi
conceded defeat.
To be sure, we can’t predict whether Modi’s acceptance of election results will
continue to mark his conduct in the future as well. Moreover, some of Modi’s
appointees to the Election Commission, the body that not only conducts the
elections but also certifies the results, continue to be civil servants, who have
been close to Modi, serving him at various points in his political career. But
that, in and of itself, is not a violation of democratic principles. Partisanship
routinely accompanies democracy, parliamentary or presidential. It is the use
of partisanship to undermine the basic processes, or the fundamental princi-
ples, of democracy that would constitute an unacceptable violation. If Modi’s
appointees to the Election Commission were to overturn the results of an elec-
tion, or deprive voters opposed to Modi of their franchise, then it would be
legitimate to claim that the fundamental tenets of an electoral democracy were
transgressed.
But that has not happened yet. Modi and his party, the BJP, not only won the
national elections of 2014 and 2019, widely viewed as free and fair, but what
is even more critical for democratic theory, they also lost key state elections
before the 2019 national victory and after. But the Modi government has not
used its power, in a constitutionally unauthorized way, to annul any of the state
elections.
To say that India is not an electoral democracy is therefore an overstretched
claim. Instead, the best description would be that the electoral aspects of democ-
racy notwithstanding, the basic structure of a liberal democracy, comprising
freedom of expression, independence of civil society, and minority rights, is
under severe erosion. The liberal deficits, though always present in India’s dem-
ocratic record, have widened to an alarming degree.41 And that is because of the
nature of the ideology that animates the Modi regime. The ideology of Hindu na-
tionalism departs from India’s constitutional values in two significant ways, each
hurting the liberal aspects of democracy.
First, for Hindu nationalists, India is a Hindu nation, an idea the Constitution
does not endorse. Demographically, Hindus are roughly 80 percent of the na-
tion and Muslims, at just a little over 14 percent, are the largest minority. Given
India’s history, Muslims are a special object of Hindu nationalist ire. The Hindu
nationalists have always viewed the Muslim community, with a few individual
exceptions, as insufficiently patriotic, or even entirely disloyal to India. Hence
their dictum that the Muslims should not have the same rights as the Hindus,
who are the “original people” of India and must be accorded cultural and po-
litical primacy.42 In its conception of the post-independence political order,
India’s Constituent Assembly (1947–50) did not even come close to the notion
of Hindu primacy. The Constitution is rooted in the idea of religious equality
and equal rights for all citizens. If anything, it gives religious minorities man-
ifest protections from the possible emergence of a majoritarian impulse in the
polity, by guaranteeing a right to the maintenance of their religion as well as cul-
ture, including language and educational institutions, supported by state grants
if necessary.43
Second, a muscular form of nationalism also accompanies the rise of Hindu
nationalists in power. This kind of nationalism threatens not only the minorities
but also the dissenting citizenry in general. According to Hindu nationalists, lib-
eral freedoms can’t build a strong nation; only national discipline and obedience
to the state can. Modi has argued that citizen duties must be accorded prefer-
ence over citizen rights (Modi 2019), whereas the liberals and leftists, according
to Hindu nationalists, tend to celebrate only individual freedoms and minority
rights, thereby sapping national strength, unity, and resolve. According to a pop-
ular Hindu nationalist discursive trope, heavily promoted by the Modi govern-
ment, many liberals are “anti-nationals,” who should be punished by the coercive
arm of the state. Unrestrained argumentative freedoms cannot be granted to cit-
izens and civil society groups, except perhaps at the time of elections. Nation-
building ought to replace the idea of citizen freedoms, once elections are over.
Aimed at taming citizen dissent, such state-supported discourse leads to an ex-
plicit or implicit attack on the institutions that cannot perform their legitimate
roles without political and intellectual freedoms—courts, universities, civil so-
ciety, and the press. It also privileges vigilante action if the vigilante groups seek
to implement the governmental vision and harass dissenters.
India’s great political paradox is that although the constitution does not
permit abridgement of the rights of religious minorities, the push against civil
liberties does have a constitutional anchorage that Hindu nationalists can, and
do, exploit. While universal adult franchise attracted near consensus among the
political elites in the Constituent Assembly, civil liberties—a cornerstone of lib-
eral if not electoral democracy—did not. Most of the early independence elites
did not want unconstrained civil freedoms for the citizenry. According to Article
19 of the Constitution, “libel, slander, contempt of court, . . . decency or mo-
rality . . . [and] security” specified limits on freedom of speech. And the First
Amendment, going further, stated that civil liberties should not prevent the
government from making any laws that restrict such liberties “in the interests
of the security of the State, friendly relations with foreign States, public order,
etc.” Suspension of civil liberties at a time of emergency was also constitutionally
allowed.
In other words, the idea of liberal deficits was written into the Constitution.
Some governments have interpreted the restrictions on freedoms more liberally
than others; others have simply viewed them more restrictively, including a use
of such restrictions for targeting critics on grounds of public order or national
security. Courts have intervened, but not always in defense of citizens. Courts
cannot violate the constitutionally enshrined restrictions; they can only interpret
whether governments have applied them properly in a given case, if the viola-
tion is brought to litigation. For all practical purposes, the extent of permissible
freedom in India has depended on the nature of the ruling party/governments.
Since the Hindu nationalist construction of freedoms is highly restrictive and
courts have only occasionally nullified the abridgement of freedoms, India’s lib-
eral deficits have increased under Modi. The kind of power and independence
the courts exercised during 1977–2014 has been on the wane.
From Ideology to Action
The impact of Hindu nationalism on statecraft has evolved in two phases. The
first phase concerned the everyday practices of the state; and the second phase
has been about formal enactment of restrictive laws.
Under the first Modi government, India saw a new form of communal
violence—lynching. The data show a spike in lynchings after 2014, with Muslims
as the main target (Basu 2021). The perpetrators were often not caught by the
police. Modi maintained silence on lynchings, and when he did speak, the
comments were perfunctory. As if this were not enough, Modi also appointed
Yogi Adityanath to the highest office of India’s biggest state—as chief minister of
Uttar Pradesh (UP), a state equal to Brazil in population. Before this high-level
appointment, Adityanath, head of a Hindu religious order, had also created a
large vigilante organization known for its anti-Muslim mob campaigns.
The ostensible aim of lynchings is to prevent beef eating and to stop young
Muslim men from marrying Hindu women, even if their relationships are volun-
tary. But the underlying aim is quite clear. Lynchings are basically aimed at pro-
ducing a political order premised upon the idea of Hindu primacy and relegation
of Muslims to secondary citizenship. In BJP-ruled states, the police hardly catch
the lynchers and if they do, the prosecutorial cases have been weak. Several BJP
politicians have openly supported lynchers (Varshney 2019c, 75–76).
Since the return of Modi with a larger parliamentary majority in May 2019,
this project has acquired a serious legal dimension as well. The BJP’s parliamen-
tary numbers can now allow the party to underwrite new legislation. Of the
various pieces of legislation, the three that are most threatening to liberal democ-
racy bear attention.
Among the first legislative acts of the second Modi government was the
amendment of the laws related to terrorism and “public safety.” The government
now has the power to designate any individual as a terrorist (based on writings,
speeches, even possession of certain kinds of literature), and the room for judicial
appeals against such detentions has been substantially reduced. A similar legal
dictum applies to preventive detention on grounds of public safety. Essentially,
with these laws, imprisonment for political dissent is perfectly possible, some-
thing not practiced on a significant scale since the Emergency but increasingly
being practiced now.44
The change in laws about Kashmir was the second major legislative act. India
is an asymmetric federation and compared to the other states, the Constitution
had given Kashmir, the only Muslim-majority state of the country, greater au-
tonomy and powers. In August 2019, consequent upon the passing of legisla-
tion introduced by the BJP in parliament, Kashmir lost its status as a state and
was brought directly under Delhi’s rule. And using the public safety act passed
a few weeks earlier and summarized above, hundreds of Kashmiri politicians
and activists were thrown into jail. In addition, civil liberties were suspended,
making citizen protest illegal. What happened in Kashmir was a replica of
the 1975–77 Emergency, but in one state, not nationwide.45 The jailed polit-
ical leaders have mostly been released, but they spent time in jail just for being
critics of the Modi regime.
The third and final piece of legislation was an amendment to the nation’s citi-
zenship law, originally framed in 1955. Called the Citizenship Amendment Act
(CAA), the new law provides fast-track citizenship to the members of “perse-
cuted minorities,” who entered India before December 31, 2014, from Pakistan,
Bangladesh, and Afghanistan. The law not only specifies these three countries
but also lists communities it designates as persecuted minorities: Hindus, Sikhs,
Buddhists, Jains, Parsis, and Christians. It leaves out only one community: Muslim
immigrants. India’s original 1955 citizenship law drew no religious distinction
between Muslims and others.
The Minister for Home Affairs, second in command after Modi, also
announced that a National Registry of Citizens (NRC) would be created as a se-
quel to the CAA. In principle, using CAA, the NRC can render virtually stateless
a large number of Muslims, if they don’t have the documents acceptable to the
government for citizenship, even if they were born in India and have lived in the
country for decades.46 As the CAA was passed and the NRC was announced as a
future move, nationwide protests, mostly peaceful, broke out for three months,
continuing until the Covid-19 pandemic stopped large groups from getting to-
gether and protesting. Shaken by the strength and breadth of protests, the Modi
government has said it will not implement the NRC. But one can’t really be sure
that the idea will not come back at a future date.
Three other points ought to be noted, two institutional, a third concep-
tual. First, of the principal institutional mainstays of democracy after Indira
Gandhi’s Emergency (see the previous section, “Democracy’s Second Innings
and the Birth of Elite Interests”), the judiciary appears to be buckling under
executive pressure. Normally shrouded in secrecy, the executive’s push against
the judiciary became public when four of the most senior judges of the Supreme
Court gave an account of the pressures applied by the Modi government.47 The
signs of the emerging subservience of the Supreme Court are too prevalent to
be missed. The legality of preventive mass arrests in Kashmir and the constitu-
tionality of Kashmir’s diminished status have been challenged in the Supreme
Court. But even the habeas corpus cases, normally an object of instant hearing,
are still to be fully heard, let alone judged. Though some of the courts have
taken the Modi government to task for its decisions, those are generally lower
down in the judicial hierarchy, and can be overruled by the highest court in
Delhi. It is the behavior of the apex court, the Supreme Court, that is causing
great anxiety in liberal circles. The idea of executive and/or legislative sover-
eignty, relatively unconstrained by judicial scrutiny, appears to have made a
substantial comeback.
However, the role of federalism—and this is the second institutional point—
as a check on the power of the central executive remains substantial. Virtually all
non-BJP state governments took a clear stand against the CAA and NRC, pub-
licly announcing their intention not to implement them in their respective states.
Modi’s government’s withdrawal of the NRC proposal may have a great deal to
do with the realities of federalism. The Modi regime has sought to truncate the
power and authority of state governments, often interfering in their domains, but
non-BJP governments have continued to push back.
Let me finally turn to a tricky conceptual matter. Modi’s government’s anti-
Muslim legislative moves, both in the form of the CAA and the change in
Kashmir’s status, were part of the BJP’s 2019 election manifesto that guided
the party’s election campaign. Though in the voting data it is hard to parse out
how many citizens voted in favor of these specific measures in comparison to
the other issues in the campaign, the Modi regime can claim, as it has, that the
change in Kashmir’s status and the amendment in citizenship laws were both
electorally approved.48 An apparently electorally legitimated attack on the liberal
aspects of the polity has thus been launched, deepening the electoral–liberal gap.
If the national majority votes even more clearly for Hindu nationalism in the
future and the state governments, one by one, fall to the BJP in state-level victo-
ries, it is politically possible to launch an attack on the Constitution, amend its
basic commitment to religious equality, and turn Muslims legally into second-
class citizens. It will be an electorally approved collapse of liberal democracy.
That is what Hindu nationalist leaders would ideally desire, for it would show
mass support, or citizen imprimatur, for their deeply held values. India has not
approached that point yet. And it may not.
Conclusion
By way of conclusion, let me recapitulate the main arguments of this chapter.

First, India’s record as an electoral democracy is far better than its record as a
liberal democracy. Electoral vitality coexists with liberal deficits. This is in part
due to a founding ambiguity in India’s Constitution. The Constitution vigorously
supports the idea of universal franchise, but by making citizen freedoms subject
to considerations of public order, not simply national security, the Constitution
has installed a weaker notion of civil liberties. The ruling governmental regimes
have basically determined how liberally, or restrictively, the idea of civil liber-
ties would be interpreted and executed. The courts have not been a consistently
strong exponent of civil freedoms.
Second, my argument about India’s democratic longevity has centered on elite
choices, not structural conditions. Elite choices, in turn, are a composite of two
different aspects: values and interests. Reflected in the Constitution, values of the
first generation of political elites accounted for the early institutionalization of
democracy. However, elite values alone do not explain India’s democratic lon-
gevity, as substantial elite interests over the last few decades have come to be as-
sociated with democracy. Power is no longer the monopoly of one party, and a
variety of parties hold power at different levels of the polity. The multiplicity of
parties and elites with a stake in power contribute to democratic perseverance.
Those parties and political actors who would be hurt by the suspension of de-
mocracy have tended to fight for their interests.
Third, the electoral–liberal gap widens when Hindu nationalists come to
power, as is true today. This happens because Hindu nationalists privilege Hindu
supremacy over the protection of minority rights and liberal freedoms. The
longer the Hindu nationalists remain in power, the weaker will India become as
a liberal democracy. But whether it will also cease to be an electoral democracy
remains uncertain.
Notes
* For comments on an earlier draft, I am grateful to Nancy Bermeo, Jennifer Bussell,

Niraja Jayal, Madhav Khosla, Sudhir Krishnaswamy, Pratap Mehta, Vishnu Pad, Dan
Slater, Arun Thiruvengadam, Steven Wilkinson, Deborah Yashar, and especially
Scott Mainwaring and Tarek Masoud. Bhanu Joshi’s research assistance is highly
appreciated.
1. Further explanations of these terms are available in Haggard and Kaufmann (2020)
and Levitsky and Ziblatt (2018).
2. It should, however, be noted that Przeworski et al.’s conception of democracy relies on only
one of Dahl’s two criteria. It is only about contestation, not participation (2000, 34–35).
3. It is noteworthy that for a democracy theorist like Przeworski, such deficits do

not constitute a democratic disqualification so long as the criterion of com-
petitive elections is satisfied. And for him, governmental turnovers are a key
indicator of electoral competitiveness. The basic intuition here is that a non-
democratic system would not allow incumbents to lose power, even if elections
went against them.
4. Also relevant, in a comparative sense, is the fall of democracy in Chile (1932–73),
Uruguay (1942–73), and Venezuela (1958–2000s). While 1932, 1942, and 1958 were
not founding moments for these countries, as 1947 was for India, there was a collapse
of democracy after decades of existence.
5. On the inclusion of freedom of expression and association under electoral, as op-
posed to liberal, democracy, see my comment in note 41.
6. Why Latin America and the Caribbean developed such a large gap after 1990–99 is
interesting, but it need not detain us here. My focus is on India.
7. In the 2019 campaign for national elections, one of the key slogans of the opposition
Congress Party was that Prime Minister Modi was a “thief.” The leader of the op-
position Congress Party believed that though Modi fought the 2014 elections on an
anti-corruption platform, a defense deal he struck in France, after coming to power,
unmistakably revealed corruption. The campaign went on for months. No one from
the Congress Party was arrested during the campaign.
8. At the national level, the winners were poorer in 1977, 1989, 1996, 1998, and 2004.
At the state level, the numbers are much greater, including the widely noted 2021
elections in West Bengal. A state-level party, Trinamool Congress (TMC), defeated
the BJP, which outspent its opponents by a huge margin but lost decisively.
9. One more exception is for some years between the early 1950s and early 1960s,
when India’s index exceeded that for Western Europe and North America. Greater
disaggregation of V-Dem data, not presented graphically here, shows that North
America pulls the average down in those years. While the indices for Canada and
the US are consistently higher than those for India, following the United Nations
Statistics Division (UNSD) methodology, V-Dem also includes in the region of
North America: Bermuda, Greenland, Saint Pierre, and Miquelon and Antarctica.
It is not clear that Greenland should be counted separately because it is not an inde-
pendent country. And Bermuda gained independence only in 1995. Because of these
complications, it is perhaps not advisable to make too much out of India having a
higher index than Western Europe and North America.
10. Two things should be added here. First, as noted earlier, for their definition of democ-
racy, Przeworski et al. (2000) do not use the full Dahlian measures, concentrating
only on contestation (not on participation, let alone the liberal freedoms). Second,
their exceptional list also includes Belize, Solomon Islands, and Vanuatu. For ease
of exposition, I only pick four from their exceptional list: Jamaica, Mauritius, Papua
New Guinea, and India.
11. These numbers are based on the World Bank’s open data site (data.worldbank.org).
In principle, one should also look at per capita incomes at an earlier point. The World
Bank gives us the following numbers at current prices for 1960: Botswana ($58),
Costa Rica ($381), Jamaica ($429), Papua New Guinea ($115), and India ($81). Only
Botswana had a lower GDP per capita then, which should, in principle, complicate
the assessment of Przeworski et al. The 1960 income statistics are not available for
Mauritius.
12. The World Bank’s middle income category ranges from a per capita annual income of
$1,026–$12,375 (at 2018 prices). India entered the lower middle-income category in
2005. China had done so roughly a decade earlier.
13. The classics on how democracy and caste could coexist are still relevant. See Rudolph
and Rudolph (1967) and Kothari (1970a).
14. In government schools, where a lot of Indians have been educated, pupils would
be taught in their mother tongue and would additionally learn Hindi and English.
Depending on the resources of a state, two languages were definitely taught virtually
all over India, and attempts for a third were also made.
15. Lijphart (1996), however, argues that India is a consociational democracy, a claim not
accepted by India specialists. The most systematic critique has come from Wilkinson,
who flatly argues that “Lijphart miscodes post-independence India as consociational”
(2004, 154).
16. Wilkinson (2004) argues that pre-independence India was consociational.
17. The quotas that were instituted were caste-based, not religious.
18. Lijphart (1999) puts India in the middle of thirty democracies in terms of the pro-
pensity to form coalition governments. Coalitions per se should not be viewed as a
core feature of consociationalism. A key question is: do coalitions mirror the diverse
groups in society? Only in recent years have India’s coalition governments at the na-
tional level incorporated a lot of lower caste politicians. The upper castes have always
had more ministers in the cabinet than their weight in the population.
19. Sri Lanka may be among the few exceptions. Its universal franchise was born in 1931,
when the British were still the rulers.
20. The exception was the province of Punjab, where the Unionist Party triumphed
in 1937. Elsewhere, the parties representing landlords could not win provincial
elections.
21. In the 1950s, too, a party representing the interests of the propertied emerged.
Called the Swatantra Party, it fought elections. But it played no role at the time of
Constitution making in the 1940s, when democratic institutions were put in place.
Moreover, it ceased to be a force after two elections. It could not successfully fight
against the Congress Party, the institutional carrier of anticolonial nationalist
sentiment.
22. In the case of the indigenous princes, the accommodation of interests worked differ-
ently. The British had sovereign powers over the entire country, but they left Indian
princes to administer their own territories, constituting a third of India. The Congress
Party steadfastly opposed the power and authority of the princes, but offered them
an olive branch: a government-funded allowance to the princes, called privy purse,
would be granted in return for the formal accession of their territories to independent
India. Almost all princes accepted the arrangement, and those very few who did not
were militarily crushed. Privy purses were a small compensation for the authority and
powers to tax that the princes lost, but they understood that the winds of change and
legitimacy were with the Congress Party.
23. In her comparison of India and Pakistan after 1947, Tudor (2013) also argues that
power was in the hands of middle-class leaders in India, not the landlords, a class that
dominated western Pakistan.
24. Poverty rates were more reliably calculated after 1961.
25. The commitment was made by the Nehru Committee (1928), which worked, among
other things, on the kind of polity India should have.
26. The legislature was split into “no less than thirteen communal and functional
compartments for whose representatives seats were reserved” (Austin 1966, 144).
27. Nehru’s letter to chief ministers, March 21, 1951. Quoted in Khosla 2014, 115.
28. Syama Prasad Mukherji was among the most prominent Hindu nationalists to be
a member of the Constituent Assembly, as was the Communist politician Somnath
Lahiri. Congress Party members, however, were roughly 80 percent of the Assembly.
29. Ambedkar was not only the most educated Dalit in history until then, but also the
most educated leader across the entire political spectrum. His first PhD, combining
law and political theory, was from Columbia University, and the second, in eco-
nomics, from the London School of Economics.
30. Ambedkar made this argument as early as the late 1920s, an argument he never
changed.
31. See the summary of the debate in Austin (1966, 46–49).
32. Here is a description of the scale of the exercise: “At stake were 4,500 seats—about 500
for Parliament and the rest for the provincial assemblies. Some 224,000 polling booths
were constructed and equipped with two million secret ballot boxes, requiring 8,200
tons of steel. To type and collate the electoral rolls by constituencies, 16,500 clerks
were appointed on six-month contracts. About 180,000 reams of paper were used
for printing the rolls. To supervise the voting, 56,000 presiding officers were chosen.
They were aided by 280,000 lesser staff members; and 224,000 policemen were put on
duty to stop violence and intimidation” (Guha 2007, 144). In addition, Shani (2018)
gives a detailed account of how the huge electoral rolls were prepared, and how the
elections took place.
33. All of this was primarily done though the 42nd Constitutional Amendment, passed
without the presence of most opposition leaders in parliament. Prakash (2018)
provides a detailed account of the Emergency.
34. This was the 4–1 decision of the Supreme Court in ADM Jabalpur v. Shivkant Shukla.
35. She did bounce back in the next elections in 1980, but she obviously did not know
that in 1977, when she lost.
36. The divergent constitutional histories of Pakistan and India are perhaps worth noting.
Carved out from the same British colony which also led to the birth of independent
India, Pakistan has witnessed three constitutions since its birth. Indeed, it could not
finalize a constitution for the first decade of its existence, by which time India not
only had a constitution in place, but also had conducted two national and many state
elections. See Tudor (2013).
37. For how he defines “consolidated democracies,” see Przeworski (2019, 29–32).
38. See especially Linz (1990), but also the critique of Linz by Mainwaring and Shugart
(1997).
39. Rajamani and Sengupta (2010) compare the pre-and post-emergency history of judi-
cial review.
40. For more details, see Divan (2016).
41. Based on V-Dem measurements, Figure 2.2 (in the first section of this chapter)
does not show the widening gap between the electoral and liberal dimensions since
2014. It suggests India’s decline on both fronts. Moreover, the descent looks sym-
metrical. It turns out that this symmetrical decline is, in fact, an artifact of the way
V-Dem measures electoral and liberal democracy. V-Dem’s electoral democracy
score is not simply based on Robert Dahl’s two minimal criteria—contestation and
participation—which are widely viewed as the best ways to measure the electoral core
of modern democracy. V-Dem’s electoral score also includes “freedom of expres-
sion . . . and association” (Coppedge et al. 2020a, 35). And its liberal democracy index,
in addition to the electoral democracy, takes note of “the importance of protecting
individual and minority rights” (Coppedge et al. 2020a, 43). The concept of liberalism
in modern liberal theory certainly includes individual and minority rights, but also
freedom of expression and association (Ryan 2012). It is unclear how to conceptu-
ally separate freedom of expression from individual civil rights. Similarly, to include
freedom of association under electoral democracy raises conceptual awkwardness,
for it includes both freedom to form and operate political parties and freedom for
non-party civil society organizations. An electoral democracy may privilege freedom
for political parties over freedom for civil society, as has become true of India under
Modi. Generally speaking, if a democracy provides freedoms both to parties and to
civil society, it is a deeper democracy, or a deeper polyarchy, as Dahl would put it, but
it is best to stick to a minimal notion of electoral democracy and a more comprehen-
sive notion of liberal democracy.
42. For details, see Varshney 2002, 2013, 2019b.
43. See Agnes 2016; Reddy 2016.
44. The V-Dem reports it as follows: “The Modi-led government in India has used laws
on sedition, defamation, and counterterrorism to silence critics. For example, over
7,000 people have been charged with sedition after the BJP assumed power and most
of the accused are critics of the ruling party” (V-Dem 2021, 20).
45. For more details, see Mehta 2019.
46. For a definitive analysis of the original citizenship law, see Jayal (2013). For details of
the new changes, see Varshney (2019a).
47. This is discussed at some length in Varshney 2019b.
48. See the analyses in Varshney (2019c) and Jaffrelot and Vernier (2020).
3
The Politics of Permanent Pitfalls
Historical Inheritances and Indonesia’s
Democratic Survival
Dan Slater
To be a new democracy is to be confronted by a plethora of pitfalls. When de-

mocracy is born in a “hard place” (Chapter 1, this volume), these pitfalls are
especially deep and dauntingly numerous. These risks are especially acute
during democracy’s fragile first days, when nobody knows for certain how the
new system is going to work, how their rivals are going to behave, and whether
they will be able to thrive under an unfamiliar and fluid set of rules (Lupu and
Riedl 2013).
This does not mean that democracy’s many pitfalls simply disappear over
time. If widespread processes of democratic backsliding and erosion across the
world throughout the difficult decade of the 2010s have taught us anything, it is
that the classic notion of democratic consolidation—democratic systems gradu-
ally becoming bulletproof against autocratic actors and impulses, especially once
they reach some magical threshold of economic development—has always been
a complacent conceit at best, and a dangerous illusion at worst.
Casting aside the idea that democracy can ever truly consolidate, I argue in
this chapter that democratic survival entails navigating a politics of permanent
pitfalls. I argue further that, to be adequately understood, this politics must al-
ways be placed in historical perspective, grounded deeply within the case where
democracy is struggling to take root and survive. My argument is not that his-
torical forces matter more than proximate factors—or pivotal actors—for dem-
ocratic survival. It is that any account of democratic survival which ignores
historical forces specific to the country in question is necessarily—in fact
woefully—incomplete.
The same can be said for any account that fails to distinguish, conceptually
and theoretically, among the myriad ways that democracy can backslide and
break down. Merely historicizing any country’s relative democratic success with
piles and piles of historical material is insufficient as well. What is needed is not
just historicizing, and not just theorizing, but historically theorizing democratic
survival. This requires, in turn, aligning the inheritances a country possesses with
the pitfalls it confronts. We cannot even know which pitfalls pose the greatest
threat to democracy in a particular country in the first place unless we first know
how historical inheritances have made some pitfalls more ominous and perma-
nent than others.
This chapter focuses on Indonesia: one of the very hardest places, and thus
one of the biggest surprises, of any case where democracy has arisen and, for the
most part, succeeded in the first two decades of the twenty-first century. How
have Indonesia’s historical inheritances helped it survive the politics of permanent
pitfalls, at least thus far?
My argument centers on two historical inheritances Indonesia enjoyed when
transitioning to democracy, in what appeared to be the hardest imaginable
circumstances, in 1998–99: (1) egalitarian nationalism inherited from before the
authoritarian period, and (2) institutional strengths inherited from the author-
itarian period itself. I argue that these two historical inheritances help explain
why Indonesian democracy has succeeded to the considerable extent that it has,
because of how they align with four major pitfalls that new democracies so often
confront: (1) state failure, (2) military takeover, (3) electoral authoritarianism,
and (4) illiberal democracy.
My definitions of these concepts draw directly from their standard usage in
the field. State failure simply means a collapse of central authority so severe that
a country becomes better described as anarchic than either democratic or au-
tocratic. Military takeover simply means the seizure of direct power by men in
uniform, as transpired in February 2021 in Myanmar. The more subtle and sig-
nificant distinctions lie between electoral authoritarianism (Schedler 2013) and
illiberal democracy (Zakaria 1997). Most pithily put, electoral authoritarianism
sees a dominant ruling group doing whatever it takes to win elections, while illib-
eral democracy sees a dominant chief executive doing whatever they want after
winning elections (Slater 2018a).
All four of these threats pose permanent pitfalls to democracy, especially in a
challenging context like Indonesia. Not all pitfalls represent equivalent threats
in any particular case, or even in general, across the world, however. I argue that,
of all four pitfalls, illiberal democracy is the hardest to eliminate, because it is the
easiest for a single irresponsible elected politician to bring about. Mainwaring and
Masoud are absolutely right to argue in this volume’s introduction that democ-
racies survive or collapse through the concrete actions of leading politicians, and
not just as a reflection of underlying structural conditions. The key point here
is that one aspiring autocratic actor can inflict illiberal democracy on a country
after gaining executive power. Building electoral authoritarianism requires a
level of elite collective action and institutional engineering that practicing illib-
eral democracy does not.
The Politics of Permanent Pitfalls 75
This does not mean that illiberal democrats can rule without consider-
able public support and elite compliance. No chief executive is an island. What
it means is that electing a single illiberal democratic leader—a Bolsonaro, a
Duterte, a Trump—can be sufficient to propel a country into a bout of illiberal
democracy, but not to establish a system of electoral authoritarianism. Actors
are absolutely critical in shaping regime outcomes; but they can push democracy
more easily into some pitfalls than others.
Like so many countries—perhaps even all countries—Indonesia has not en-
tirely escaped, and can likely never entirely escape, the pitfall of illiberal democ-
racy. Whether Indonesia will avoid succumbing to an illiberal democratic fate
depends largely on whether its deep but increasingly distant inheritance of egali-
tarian nationalism can withstand the challenges of those who would undermine
and undo it.1 Indeed, it has been Indonesia’s lingering vulnerability to illiberal
democracy that explains the recent softening in the country’s democracy ratings.
But it is vital to appreciate that the other three main threats to democracy—state
failure, military takeover, and electoral authoritarianism—have all been success-
fully evaded. Those critical escapes are likely to continue, and these are no mean
accomplishments.
In the following section I detail why Indonesia qualifies as a hard place for
democratization. It was not just the country’s underlying features, but also the
disastrous environment in which democratic transition occurred, that should il-
lustrate why Indonesia’s democratic success was emphatically against the odds.
The subsequent two sections explain why the egalitarian nationalism that arose
during Indonesia’s fierce independence struggle and the institutional strengths
that emerged under Suharto’s authoritarian New Order regime (1966–98) have
given Indonesia a set of historical inheritances that have helped it beat those long
odds. The chapter’s fourth section distinguishes four common pitfalls that new
democracies face—state failure, military takeover, electoral authoritarianism,
and illiberal democracy—and details how Indonesia’s historical inheritances
have served as resources for avoiding each of those pitfalls.2 I pay particular at-
tention to the significance of egalitarian nationalism in fending off the massive
threat of right-wing populist-authoritarianism that Indonesia confronted in the
presidential elections of both 2014 and 2019. The conclusion remarks on the en-
during and even endemic danger of illiberal democracy—the most permanent of
pitfalls—not just in Indonesia but across the democratic world.
Indonesia as a Hard Place
There are at least seven reasons in the comparative politics literature why
a country might be considered a hard place for democracy to emerge and
thrive: (1) economic underdevelopment, (2) economic crisis, (3) ethnoreligious

polarization, (4) Muslim majority, (5) large population, (6) authoritarian neigh-
borhood, and (7) national disunity. Without arguing that all of these factors are
indeed barriers to democratization, the fact remains that they are commonly seen
as barriers, and that as of 1998 Indonesia arguably confronted all seven of them—
some of them in spades.
We should begin with the obvious. Indonesia is the fourth most populous
country (factor #5) and has the largest Muslim population in the world (factor
#4). To the extent that democracy correlates with countries where both the total
population and the Muslim population are smaller—weakly defensible causal
claims but persistently robust correlations nonetheless—Indonesia democra-
tized with a double demographic difficulty. Southeast Asia has also always been
a predominantly authoritarian regional neighborhood (factor #6), with a super-
majority of countries being stably authoritarian rather than even intermittently
democratic. As of 1998, when Indonesia commenced its democratic transition,
only the Philippines and Thailand qualified as electoral democracies in Southeast
Asia, and neither could even remotely be said to be the kind of model democ-
racies that might provide a positive demonstration effect. Although Indonesia’s
close relations with the United States, as forged during the Suharto era, arguably
made democratization more likely both to arise and endure, there were certainly
no regional tailwinds aiding Indonesia’s regime transition—especially when
one adopts a wider regional lens and considers the growing influence of China
along Asia’s Pacific rim. Indonesia also clearly confronted the stiff headwinds of
national disunity (factor #7), as the restive provinces of Aceh, East Timor, and
West Papua all credibly threatened to use their new democratic rights to separate
from the Indonesian republic. In the case of East Timor, this separation indeed
occurred, first through the peaceful means of a referendum, and then violently as
the Indonesian military and its local militias exacted a horrible price on the East
Timorese for decisively rejecting Jakarta’s harsh embrace.3
The first three barriers mentioned above require a bit of deeper discussion.
Especially complicated is the question of ethnoreligious polarization (factor
#3). To be sure, Indonesia is one of the most ethnically, linguistically, and reli-
giously diverse countries in the world. The country has no shortage of cleavages
that have manifested into conflict at one time or another, including Muslim and
non-Muslim, Javanese and non-Javanese, “indigenous” (pribumi) and “Chinese,”
and, at the local level, all manner of tensions between majority and minority
ethnic groups (Slater 2015). Yet, ironically, Indonesia’s sheer number of iden-
tity cleavages makes it difficult for any single point of friction to combust into
national flames. Horowitz (2013) has gone so far as to argue that Indonesia’s
complex cleavage structure is more of a benefit for democracy and stability
than a threat. Perhaps the most judicious way to put it is that Indonesia’s riotous
diversity presents both challenges and opportunities for crafting a stable demo-
cratic and pluralistic order. But surely the fact that Indonesia lacks the relative
ethnic homogeneity of a Japan or South Korea makes it more of a “hard place” for
democracy than it would otherwise be.
This brings us to the first two theorized hindrances to democratization men-
tioned above: economic underdevelopment (factor #1) and economic crisis
(factor #2). The latter is especially significant in the Indonesian case. Among all
the countries that have democratized since the third wave began in the 1970s,
none has done so amid more dramatically worsening economic conditions than
Indonesia. As Figure 3.1 indicates, the drop in economic growth rates from the
start of the Asian financial crisis in 1997 until its bottom in 1999 was truly precip-
itous. This plunge in real economic output was paralleled in the financial sector,
where Indonesia’s currency (the rupiah, or Rp) plummeted from over Rp2000/$
to under Rp13000/$ in a matter of weeks, and the private banking sector utterly
collapsed under the weight of its dollar-denominated debts. To the extent that
democratizing during economic crisis produces rockier outcomes than democ-
ratizing in good economic times (Haggard and Kaufman 1995), the conditions
surrounding Indonesia’s democratic transition could hardly have been less
auspicious.
Yet this is not to say that Indonesia democratized without any economic
strengths to its name. As Figure 3.2 indicates, Indonesia’s fall was so dramatic in
15
10
5
GDP Growth (annual %)
–5
–10
–15
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
Figure 3.1. Indonesian economic growth rates, 1961–2019

Source: World Development Indicators n.d. (indicator: NY.GDP.MKTP.KD.ZG)
5,000
4,500
4,000
GDP per capita (constant 2010 US$)
3,500
3,000
2,500
2,000
1998
US$ 2084.1
1,500
1,000
500
0
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020
Figure 3.2. Indonesian economic development levels, 1960–2019

Source: World Development Indicators n.d. (indicator: NY.GDP.PCAP.KD)
large measure because its rise had been so stratospheric. Suharto’s authoritarian
rule from 1966–98 may have been an era of colossal corruption, but it was also
unmistakably an extended period of rapid and sustained growth.
When Suharto seized power in a counterrevolutionary bloodbath against the
Indonesian Communist Party (PKI) and its suspected sympathizers in 1965–
66, the country was an economic basket case. Hyperinflation and hunger ran
rampant from the late 1950s through the mid-1960s, as Indonesia’s founding
father and president, Sukarno, miserably mismanaged the economy during
the country’s years of “Guided Democracy” (1959–65). The economy was
overwhelmingly agricultural and far from self-sufficient in rice, the nation’s
staple crop.
By virtue of massive influxes of foreign aid, oil revenue, and economic reforms
designed to generate new investment both foreign and domestic, the Indonesian
economy had made a dramatic turnaround by the mid-1970s. Far from catering
to urban dwellers at the expense of the countryside, Indonesia’s economic
policymakers made both rice self-sufficiency and family planning their highest
priorities, easing the rural misery that had helped galvanize the PKI’s rise into
one of the largest communist parties in the world. Economic growth then spiked
considerably in the 1980s, after the US-Japan Plaza Accords strengthened the
yen and sent avalanches of Japanese foreign direct investment into Indonesia and
its low-wage industrializing neighbors.
The upshot was that by the late 1990s, Indonesia’s GDP per capita had
surpassed $2,000 (constant 2010 US$; World Development Indicators, n.d.),
and urban areas were increasingly filled with educated middle-class citizens
who resented the Suharto regime’s astonishing corruption and rankled under
its smothering blanket of authoritarian controls. It would be from these urban
middle-class sectors that the bulk of protesters would come as the Suharto re-
gime floundered in the face of financial meltdown.
Suharto would step down peacefully and constitutionally in May 1998,
handing over power to his handpicked Vice President B. J. Habibie. Under con-
tinuing pressure from Jakarta’s civil society, Habibie would quickly take the lead
to remove authoritarian controls and accelerate Indonesia’s electoral timetable to
schedule free and fair parliamentary elections in June 1999. By the time Habibie
was removed in a parliamentary vote and a new democratically constituted gov-
ernment was installed in October 1999, Indonesia’s democratic transition could
be said to be complete.
In the pages that follow, much more will be said about the strengths and
weaknesses of Indonesian democracy. For now, it is simply essential to establish
that Indonesia has, indeed, remained a democracy by standard global measures
from 1999 until the present day. As Figure 3.3 indicates, Indonesia’s democracy
score shot upward with the transition of the late 1990s, improved still further in
the first decade of the 2000s (even as Southeast Asia’s other electoral democra-
cies, the Philippines and Thailand, suffered either serious backsliding or outright
collapse), and softened only slightly throughout the 2010s.
0.6
0.5
0.4
0.3
0.2
0.1
0
1953
1956
1959
1962
1965
1968
1971
1974
1977
1980
1983
1986
1989
1992
1995
1998
2001
2004
2007
2010
2013
2016
2019
Figure 3.3. Indonesia’s liberal democracy score over time, 1953–2019

As I discuss below, it has been Indonesia’s lingering vulnerability to illiberal

democracy that explains the softness in the country’s democracy rating. The
other three main threats to a young democracy in a hard place like Indonesia—
state failure, military takeover, and electoral authoritarianism— have been
successfully evaded, at least thus far. To help us understand why, the next two
sections explore the two sources of historical inherited strengths—(1) egalitarian
nationalism, and (2) authoritarian institutional development—that underpin
Indonesia’s surprising democratic survival and impressive levels of democratic
success.
Egalitarian Nationalism
Nationalism has a bad rap in liberal circles, and not without reason. Treating
the nation as a higher priority than more universal values can lead to harmfully
prideful behaviors abroad and dangerously prejudicial actions at home. For pre-
sent purposes, the critical point is that nationalism is often deployed against
ethnic and religious minorities, defining them as second-class citizens unde-
serving of the full liberal protections that are supposed to accompany democracy
as it is substantively understood. For democracy to thrive, it may seem, nation-
alism needs to be tamed.
Yet nationalism comes in many different varieties, and not all of them work at
cross-purposes to democratic development. Two main distinctions are impor-
tant.4 First, the national political community can be defined in more inclusive
terms or exclusive terms when it comes to ascriptive identities such as ethnicity
and religion. As scholars have long argued, nationalism can be more civic or
more ethnic in character, and this matters greatly for democratic development.
Second, some nations are built through the mobilization and lionization of or-
dinary people, whereas others are born without removing traditional, heredi-
tary, and feudalistic elites from their comfortable perches as fathers (and almost
never, seemingly, mothers) of the nation. One might call this the distinction be-
tween elitist nationalism and popular nationalism (Vom Hau 2008).
When nationalism is both civic and popular in character, it can be consid-
ered egalitarian, whereas nationalisms that combine ethnic and elitist features
might be called hierarchical. The first main argument I make in this chapter is
that Indonesian nationalism is far more egalitarian than hierarchical, especially
in comparison to most of its South and Southeast Asian neighbors, and that this
has served as one of the most significant historical contributors to Indonesia’s
substantial democratic success.
Indonesia’s egalitarian nationalism was a product of its revolutionary path to
independence in 1949. As by far the most significant and valuable colony in the
Dutch empire, Indonesia was not going to gain its freedom without a fight. If
India was the jewel in Britain’s crown, Indonesia was Holland’s entire diamond-
encrusted tiara. So when World War II ended and Japan surrendered to the Allies
after its four-year reign over Indonesia, the Dutch returned to the scene with
every intention of staying for good.
What followed was a bloody and brutal battle for independence, known
simply in Indonesian nationalist lore as perjuangan: struggle. The most promi-
nent figure in this struggle was Sukarno, a charismatic, polyglottal orator who
had been inspiring popular mobilization against the Dutch, across every imagi-
nable social divide, for decades. Sukarno’s consistent message was that Marxists,
Muslims, and nationalists could all unite, and indeed must unite, for Indonesia
to expel the Dutch. All good things would follow from ending colonialism.
Sukarno spent much of the prewar period in a colonial prison for expressing
these ideas and mobilizing them as the leading founder of the Indonesian
Nationalist Party, or PNI. Japan’s conquest of Indonesia was a godsend for
Sukarno and the nationalist movement he led, since Japan aimed to rule with
the collaboration of popular anti-Dutch nationalists rather than traditional
pro-Dutch collaborators. They even went so far as to mobilize an entirely new
Indonesian army dedicated to national self-defense.
Under Japanese rule, egalitarian nationalists like Sukarno shared power,
not prison cells. That left them in an ideal position to lead the violent popular
struggle for independence against the returning Dutch, and to make their ideals
the founding principles of the newborn republic.
These nationalist ideals were neither exclusionary nor elitist. In both its ci-
vilian and military wings, Indonesia’s nationalist movement was led by fig-
ures lacking hereditary status or any other kind of hierarchical standing above
common people, also known as the rakyat, a word that gained mythic nation-
alist power rivaling both perjuangan (struggle) and merdeka (independence).
Perhaps even more significantly, Indonesia’s leading nationalists fended off
efforts to define the nation in religious terms, either by making the majority reli-
gion of Islam a recognized national religion or by mandating that Muslims must
follow Islamic law. Although Indonesia would not be defined as a secular na-
tion, like India, the state would maintain at least a formal equidistance from all
of the country’s major religions. According to the official national philosophy of
Pancasila, belief in a single God is mandatory, but the practice of any particular
faith is not.
Egalitarian nationalism does not ensure democracy—no single causal factor
does, including economic development—but it does help secure it and sus-
tain it. When dictators become too corrupt, popular nationalism helps instill
the sense of popular sovereignty that draws crowds into the streets to topple
them. When forces of ethnic or religious intolerance attack minorities in ways
that are incompatible with substantive democracy, civic nationalism provides

a set of narratives, laws, and practices that can help defend disfavored groups
against those who would menacingly reduce a diverse demos to a narrow ethnos
(Mann 2004).
To be sure, types of nationalism are ideal types, and no nation is ever defined
in entirely egalitarian or hierarchical fashion. Equally importantly, the triumph
of egalitarian nationalism at the moment of independence does not eliminate
those favoring more hierarchical visions of the nation from the political scene.
Initial battles to define the nation in egalitarian terms evolve into ongoing battles
to defend the nation from those who refuse to accept that all citizens—regardless
of ethnicity or religion—should have the same rights and equal political status.
Indonesia’s egalitarian nationalism is far from pristine, and it is even further
from unchallenged, as we will explore below, when discussing the presidential
elections of 2014 and 2019 (in the section titled “Avoiding Four Pitfalls”). Yet it
provides an enduring resource for democratic development and defense that al-
most all of Indonesia’s neighbors sorely lack.
Authoritarian Institutional Development
Every democracy in the world relies, to some degree, on features that first devel-
oped before that country became a democracy. Perhaps a vibrant urban middle
class emerged during authoritarian times. Maybe a smoothly functioning legal
system was built by a colonizer with nothing but authoritarian intentions.
Sometimes a sense of national self-consciousness, solidarity, and social capital
emerges in the very process of collectively defying a tyrant, whether homegrown
or foreign in origin. Widespread literacy might have first emerged through reli-
gious instruction when a country was still ruled by a precolonial dynasty.
Of most importance for our discussion here, sometimes authoritarian
regimes build institutions that a democracy can inherit and turn to purposes of
democratic stability, even if that was by no means the original intent of those
institutions’ creators. Institutional development under Suharto’s authoritarian
regime was especially impressive in two domains, both of which ultimately paid
surprising and unintended dividends for democratic stability.
The first was state capacity. The Dutch colonial state had been world-renowned
for its governing effectiveness—albeit overwhelmingly for repressive and extrac-
tive purposes—but this institutional inheritance was largely squandered during
Indonesia’s first two decades of independence (Anderson 1983). Once an author-
itative executor of commands across the vast archipelago, Indonesia’s inherited
postcolonial bureaucracy found itself starved of resources and bereft of opera-
tional autonomy during the hyper-mobilized years of revolution, parliamentary
democracy, and Sukarno’s “Guided Democracy.” Sukarno profoundly mistrusted

political parties, but he respected bureaucrats even less. By the time Sukarno was
toppled in the confusion of conspiracy, communism, and coup d’état that ripped
the country apart in 1965–66, the Indonesian state was in no position to govern,
and there was scarcely a functioning private economy to govern in any event.
The tide turned dramatically with Suharto’s takeover. In the transition’s first
phase, anti-communist elements in the Indonesian military, led by Suharto and
allied closely with Islamic civil society, unleashed a horrific bloodbath against the
Communist PKI and its suspected sympathizers. Hundreds of thousands died;
millions were traumatized. It was upon this mountain of corpses—and it would
be a disservice to those lost souls to call it anything less—that Suharto’s “New
Order” would originally be built. Yet with the truly enormous and epically tragic
exceptions of the PKI and, after the mid-1970s, East Timor, the Suharto regime
governed less through deadly repression than through smothering coercion and
sweeping cooptation. The anti-communist mass killings were not the exception
that proved the rule; they were the exception that established New Order rule.
It was only through the wholesale reconstruction and revamping of the
Indonesian state that the Suharto regime became as stable—even, for long
stretches, boringly stable—as it was. The military was the political heart of the
New Order, but it was by no means its organizational entirety. For starters,
Suharto invested massive new resources in the bureaucracy, at first thanks to
the revenue floods of Western foreign aid and the OPEC oil boom, and later by
virtue of the booming revenues that restored economic growth and foreign di-
rect investment made possible. Military officers were typically given a variety of
leading positions in government ministries, but this was more to ensure the po-
litical loyalty of the bureaucracy than to install full-blown military governance.
Bureaucrats were generally entrusted and empowered to govern, in areas ranging
from family planning to rice self-sufficiency to managing price volatility in vital
basic commodities. Considering that bureaucrats overwhelmingly supported
the New Order as a bulwark against communism and as a welcome source of de-
velopmental and technocratic energy—in addition to the obvious risks of openly
associating either with communist or Sukarnoist ideology—Suharto need not
have worried that bureaucrats needed terribly much military oversight to do the
regime’s bidding.
The upshot of the state-building that unfolded during the three decades of
the New Order was a highly impressive track record of economic growth and
a remarkable run of relative political stability (Smith 2007). To be sure, both
growth and stability were fueled by colossal corruption that ultimately helped
bring the Suharto regime down in the 1997–98 Asian financial crisis. Yet the
fact remains that a Leviathan quite capable of governing was one of the Suharto
regime’s most important byproducts. As we discuss in the following section,
this has helped Indonesian democracy avoid the pitfalls of state failure and mil-
itary takeover, especially.
The second critical domain for institutional development under Suharto’s
New Order was that of political parties. Since the regime had such widespread
support across the political spectrum—a spectrum newly and violently shorn of
its left wing—it did not fear a return to highly controlled electoral and party pol-
itics within the first decade of its founding. Although it was a military regime, it
commanded tremendous civilian support. Party development and electoral pol-
itics provided these supportive civilians with routes to influence and largesse,
even under the suffocating coercive blanket of New Order rule.
Most significantly, Suharto quickly supported the building of a regime-
supporting political party called Golkar. The party’s name reveals both its origins
and its governing purpose. Short for “golongan karya” (functional groups),
Golkar was a political vehicle constructed from the wide variety of conservative
political organizations that arose to counter the radical leftist mass mobilization
of Sukarno and the PKI during the early to mid-1960s. In formal terms, Golkar
was not a political party but an umbrella organization—symbolically, a hov-
ering banyan tree—under which all these political organizations could shelter
and coalesce. This made it easier for Suharto to mandate that all Indonesian civil
servants become Golkar members, since the organization was more like a bu-
reaucratic superministry than a partisan vehicle in formal terms. Yet in prac-
tice, Golkar would compete in national elections every five years against the two
parties that the regime permitted to form—the Islamic PPP and the nationalist
PDI—and crush them. In essence, the state apparatus itself assumed the electoral
role of a political party, without surrendering any of the powers or resources it
held by virtue of being Indonesia’s ultimate sovereign authority.
By both hook and crook, Golkar commanded overwhelming electoral sup-
port from the New Order’s founding election of 1971 until its final election of
1997. Unlike the many pure military regimes and single-party regimes that
transitioned to democracy after the Cold War with no recent electoral experi-
ence, Indonesia became a democracy in 1999 with a firmly established electoral
system already in place. Party competition had been tightly governed, but it had
not been absent. Moderate and conservative politicians knew that elections were
something they could continue to win, at impressive levels if not at the land-
slide levels of the authoritarian era, even after full democratic competition was
installed. Golkar was the only political party with an established presence at the
local level across the entire Indonesian archipelago.
The economic crisis and surrounding corruption that felled Suharto in the
mass protests of 1998 left Golkar weakened but by no means destroyed. For all
the uncertainty that surrounded Indonesia’s democratic transition in 1998–99,
regularized democratic procedures quickly served more as a source of stability
than of instability amid the general tumult. The stoutness of the Golkar-led party
system that had been built under the Suharto regime was critical to the sur-
prising democratic stability that followed it.
At the end of the day, institutions matter because they provide predictability.
Given the extreme unpredictability of politics during times of democratic tran-
sition in “hard places”—especially when that transition is sparked by a severe
crisis—it is enormously helpful for a young democracy’s stability to have func-
tioning party and state institutions. The sad fact is that when new democracies
do not inherit such strong institutions from the authoritarian period, complete
with many of the unseemly personnel who commanded them, they either try
to build them from scratch under the most unpropitious of circumstances, or
simply try to make democracy function without them. Indonesia democratized
with a heavy institutional inheritance, but not an entirely burdensome one.
Avoiding Four Pitfalls
Explaining democratic success requires thorough and systematic attention to

the variety of ways that democracies can fail. At least four distinct pitfalls tend
to confront new democracies, and Indonesia has been no exception. A new
democracy can experience (1) state failure, if governance fails and either the
country falls apart in civil war or separatism, or ruling elites take extreme au-
thoritarian measures to prevent such outcomes from arising. It can also fall
through (2) military takeover, if men in fatigues refuse to accept the hindrances
that democratic politics presents. In less extreme but increasingly common
scenarios, democracy can backslide into (3) electoral authoritarianism, if
politicians start doing whatever they want to win elections, or (4) illiberal de-
mocracy, if they start doing whatever they want after winning them. Put oth-
erwise, electoral authoritarians target their coercion against opposition parties
seeking to topple them through the ballot box; illiberal democrats take coercive
aim at minorities, critics, and protesters who refuse to accept quietly and quies-
cently their majoritarian mandate.
In light of the arguments offered in this chapter, how have Indonesia’s
inheritances of egalitarian nationalism and institutional strength helped the
country avoid wrecking on these multiple sharp shoals? And why does illiberal
democracy continue to present the pitfall into which Indonesian democracy re-
mains likeliest to plunge?
State failure loomed especially large as a potential outcome during Indonesia’s
tumultuous transition years. The provinces of East Timor, Aceh, and West Papua
all presented credible separatist claims. Ethnic and religious violence erupted in
various parts of the archipelago, most notably in the eastern Indonesian districts
of Ambon, Maluku, and Poso. Riots against Indonesia’s ethnic Chinese minority
hit as close to home as Jakarta itself. Islamist terrorism delivered repeated blows,
most infamously in Bali. Leading country experts quite seriously asked whether
the Republic could endure without the military holding it all together by force.
Less than a decade removed from the collapse of the Soviet Union and Yugoslavia,
the notion that powerholders in Jakarta would fail to keep the country intact and
minimally governable was far from far-fetched.
Inheritances of strong nationalism and a strong state proved up to these
challenges, however. Remarkably in hindsight—and even in real time—East
Timor was permitted to exit Indonesia by popular referendum, though pro-
integration militias long backed by the Indonesian military inflicted a horrible
price on East Timor’s people for so choosing. Not only was it remarkable that
East Timor was allowed to secede; equally remarkable was the lack of separa-
tist sentiment in other parts of the archipelago which, like East Timor, had
sizable Christian majorities. Even as Christian communities feared eradica-
tion in the deadly religious conflicts that erupted in demographically divided
districts like Ambon and Maluku, no wider push for Christian separation and
self-determination gathered steam. In part, this was because national politicians
invoked the pluralistic version of nationalism, embodied by the concept of
Pancasila, in their condemnations of religious violence. Christians might not
have been made to feel at home in certain areas, but their rightful place in the
Indonesian nation was never seriously questioned.
The Indonesian state also came decisively to the rescue of the country’s
founding “unity in diversity” ideal. Steeled by over three decades of military rule,
the Indonesian military and security services were quickly deployed to conflict
zones across the archipelago, and they restored peace in surprisingly short order.
A national policy of decentralization also allowed divided provinces to split,
making local ethnoreligious conflicts less intractable. Islamist terrorism was also
effectively snuffed out by highly professionalized intelligence services. In sum,
state failure was avoided because the Indonesian nation was inclusive enough
and the Indonesian state was capable enough to prevent the Republic from vio-
lently unraveling.
While the Indonesian military was strong enough to help keep the country
together, it did not use this strength to reassume direct control of the political
system. Herein lay perhaps the biggest surprise of Indonesian democratization.
During the years preceding Suharto’s fall, the overwhelming consensus among
Indonesia-watchers was that his New Order would be followed by some variety
of collective military rule.
Yet this did not come to pass. The first reason was because Suharto had in-
creasingly personalized, factionalized, and to some extent Islamicized the mil-
itary during his final years in power. This left the Indonesian military as the
country’s most powerful actor, but not a highly cohesive actor, as Suharto exited
the scene in the face of cascading mass protests and elite defections. Of partic-
ular importance was the intense factional rivalry between Suharto’s son-in-law,
Prabowo Subianto, who commanded the military’s strategic units in and around
Jakarta, and the national head of the military, Wiranto. Prabowo proved ready
and willing to unleash violence as a way of justifying martial law and his own
assumption of dictatorial powers—much as his father-in-law had done in 1965.
Wiranto and other professional soldiers saw Prabowo as an up-jumped oppor-
tunist who was a scourge on the military rather than a savior of national stability.
It was more important to Wiranto and his ilk to restrain Prabowo and restore
military unity during crisis times than it was either to keep Suharto in power or
to salvage the military’s leading governance role.
This is not to say the military simply stepped aside. For the first five years of
Indonesian democracy, the military retained a sizable proportion of appointed
parliamentary seats, and prominent military officers played a major role in elec-
toral politics. Indeed, Indonesia’s first directly elected president, Susilo Bambang
Yudhoyono (2004–14), was one of the top military officials of the late Suharto
era. Yet the military surrendered its parliamentary seats with nary a fuss in the
constitutional revisions that followed democratization in plenty of time for the
2004 national elections.
The best explanation for why the military did so lies in Indonesia’s institu-
tional inheritance. Electoral support in 1999 and 2004 did not flow to former
radical opponents of the New Order or to proponents of root-and-branch mili-
tary reform, but to conservative and moderate parties and politicians with deep
experience in Indonesia’s military-led political system. Of particular importance
was the continued leading role played by Golkar, which has remained one of
Indonesia’s top electoral performers and has consistently secured leading roles
for itself in governing coalitions and cabinets. Although a colorful cornucopia
of new parties has emerged during Indonesia’s twenty-plus years of democracy,
most of them derive their leadership from former leading figures of Golkar. And
although PDI and PPP were designed to be mere shadow parties under Suharto’s
New Order, their post-authoritarian incarnations have thrived. To the consider-
able extent that the military can continue to rely on familiar moderate politicians
with deep New Order roots, there is simply no manifest reason for military men
to dominate the civilian arena.
Besides avoiding the fates of state failure and military takeover, Indonesian
democracy has not backslid into electoral authoritarianism. This arises when
one party or political leader gains majoritarian control through elections, then
uses that control to stifle opposition and competition for the country’s top po-
litical posts. The main reason Indonesia has been at least relatively immune to
electoral authoritarianism is its sheer scale and diversity, combined with the role
of egalitarian nationalism in ensuring that Indonesia’s vast array of ethnic and re-
ligious communities enjoy a rightful role to play in national political life.
Golkar was unable to maintain the kind of majority support that could
have allowed it to rebuild the electoral authoritarianism of the New Order era.
Electoral support in 1999 flowed to the party vehicles of egalitarian nation-
alism: the PDIP representing Indonesia’s Sukarnoist, pluralist, Pancasila ideolog-
ical stream; the PKB channeling the world’s largest Islamic social organization,
the Nahdlatul Ulama (NU); and PAN as the main vehicle for another massive
Islamic group, the Muhammadiyah. All of these parties had roots in the ideolog-
ical developments of the nationalist struggle and had maintained much of their
vibrancy during the New Order.
If any party had early prospects of capturing majority electoral support, and
at least potentially imposing electoral authoritarianism from that perch, it was
PDIP. Led by Sukarno’s daughter, Megawati Sukarnoputri, PDIP secured over
33 percent of the 1999 parliamentary vote. Yet the robustness of Islamic parties
as well as the continued kingmaking role of Golkar ensured that Megawati was
brought into sweeping power-sharing arrangements rather than being allowed
to rule alone after she was elevated to the presidency in 2001. In terms of insti-
tutional inheritance, perhaps the most vital point is that Golkar proved strong
enough to help stabilize democracy in tandem with other leading moderate
parties, but far too weak to impose a new electoral authoritarian regime on its
own (Slater and Wong 2018).
Social complexity and diversity may be a permanent challenge to democracy,
as scholars have long recognized (Rustow 1970). Less appreciated is the fact that
it also complicates the task of any would-be electoral autocrat trying to assemble
a winning coalition across cavernous spatial and demographic divides. Suharto
brought pluralists, Islamists, Javanese, and non-Javanese into his ruling author-
itarian coalition by force, with a significant assist from the binding agent of a
widely perceived communist threat. Assembling a winning electoral majority
without significant recourse to coercion is a more enormous political task in a
land of “multipolar fluidity” (Horowitz 2013) like Indonesia than in a country
where electoral appeals to a single religion or ethnicity can conflagrate nation-
wide without hitting any firewalls.
And yet electoral majorities must be assembled democratically, every five
years, in direct presidential elections. This fully democratic process could carve
a path to electoral authoritarianism in Indonesia, if and only if a majority co-
alition solidifies over time behind a president who lacks commitment to
upholding the fairness of the electoral system. With Indonesia’s two-term limit
on presidents, this task is complicated, but not impossible. The question then
becomes: has there been any prospect in Indonesia of a freely elected president
not only ignoring institutions of horizontal constraint and abusing critics and
minorities (i.e., illiberal democracy: more on which below), but also assembling
a steady majority coalition dedicated to tipping the electoral playing field in its
favor and keeping a minority opposition coalition out of office in perpetuity?5
Without question, the clearest threat to establish electoral authoritarianism
in Indonesia in such a fashion has arisen via the reemergence of Prabowo
Subianto and his populist party vehicle, Gerindra. After being dishonorably
discharged from the military and spending a spell overseas burnishing his eco-
nomic fortunes if not his political reputation, Prabowo returned to the scene and
attempted to win (in other words, buy) the chairmanship of Golkar. Defeated in
that effort, Prabowo founded Gerindra and garnered a spot as Megawati’s run-
ning mate for the vice presidency in 2009. The Megawati-Prabowo team was
trounced in President Yudhoyono’s landslide re-election. Prabowo then rose to
the top of the ticket in both 2014 and 2019, standing as one of the two presidential
candidates in both contests. Both times Prabowo would face off against former
Jakarta Governor Joko Widodo (Jokowi), a member of PDIP and a seeming stal-
wart of Indonesia’s pluralist, Pancasila tradition.
The stage was set, both in 2014 and 2019, for the onset of electoral author-
itarianism. All it would require was a Prabowo victory. To understand why, it
is essential to understand how party coalitions work in Indonesia. Despite the
country’s “multipolar fluidity” (Horowitz 2013)—or perhaps in some respects
because of it—presidential elections are always followed by the construction of
vast, supermajority coalitions transcending most if not all key identity divides
(Slater 2018b). When PDIP leader Megawati was president from 2001 to 2004,
every party except a tiny Islamic party joined her government. When retired
general Yudhoyono was president from 2004 to 2014, every single party except
PDIP shared power. If Prabowo had won in either 2014 or 2019, this coalitional
arrangement surely would have been reproduced. Even though Gerindra would
not have majority support on its own, every party except PDIP would have
backed Prabowo after he won the presidency. This would have given him the
sweeping legislative majority necessary to attack Indonesia’s voting system and
keep the lone remaining opposition party, PDIP, out of power by means foul
rather than fair.
Indonesia’s political elites were thus in no position to stop a Prabowo-led
march into electoral authoritarianism. But Indonesia’s voters were. And in both
2014 and 2019, Prabowo was soundly defeated by Jokowi, by approximately a
55–45 percent margin. It is impossible to say with absolute certainty whether
Prabowo would have attempted to undermine Indonesian democracy to the
point that it would be better considered electoral authoritarian, or would have
succeeded at doing so. What is certain is that he has simply lacked the numbers—
more specifically, the voters—to give it a try. Indonesian democracy has thus
passed its sternest tests by defeating a likely aspiring authoritarian at the ballot
box, rather than desperately trying to restrain him after he has gained presiden-
tial powers.
This leaves illiberal democracy as the main lingering threat to Indonesian de-
mocracy. Whereas electoral authoritarianism would see an elected leader stifling
political opposition, illiberal democracy in Indonesia would most likely see him
attacking religious minorities as second-class citizens. The very diversity that
tempers the threat of electoral authoritarianism, as just discussed, simultaneously
and permanently tempts opportunistic politicians to become illiberal democrats.
It is here that egalitarian nationalism has mattered most, and must continue
to hold the line against those who would undermine it. In both the 2014 and
2019 presidential campaigns, Prabowo began assembling an alliance of conser-
vative Muslim forces for his battles against the PDIP and Jokowi. This pattern
had already begun to emerge in the 2004 and 2009 campaigns, when Yudhoyono
curried favor with conservative Islamists in his successful efforts to trounce
Megawati and her more pluralist electoral coalitions. Yet Prabowo raised the
religious sectarianism to another level. This increasing religious polarization
showed its greatest effect in 2017 when ethnic Chinese Jakarta Governor Basuki
Tjahaja Purnama, popularly known as Ahok—the folksy and plain-spoken run-
ning mate for Jokowi in his 2012 gubernatorial campaign—was defeated in his
re-election campaign and subsequently imprisoned for insulting Islam in the
wake of massive protests to bring him to justice for doing so. The defeat and im-
prisonment of Ahok sent chills down the spine of every believer in Indonesia’s
ethnic and religious pluralism. The looming question in the wake of Ahok’s
jailing was whether the Islamic mass movement that arose to bring him down
would then succeed in 2019 by doing the same to Jokowi, removing a president
from power who hews to pluralist nationalism rather than Islamism.
Jokowi prevailed in that contest, but only by tilting further in the direction
of illiberal democracy than his pluralist backers preferred. Most importantly,
Jokowi replaced old Golkar stalwart Jusuf Kalla as his running mate, instead
selecting a deeply conservative Islamic leader from the NU, Ma’ruf Amin, to be-
come his new vice president. The clear goal was to inoculate himself from reli-
gious attacks without squandering his overwhelming support among pluralist
nationalists. Although the Islamist rhetoric was even stronger in 2019 than in
2014, it was almost surely more muted than it would have been if Jokowi had
not protected his Islamic flank as he did with his controversial choice of run-
ning mate.
The best analytical post mortem on the 2019 elections makes the pluralist
foundations of Jokowi’s big re-election win abundantly clear (Pepinsky 2019).
In the starkest terms, Jokowi’s victory was an enormous win for non-Muslims
and their continued status as full citizens in the Indonesian nation. Jokowi’s
stupendous advantage over Prabowo among non-Muslims became even more
pronounced in 2019 than in 2014. In Bali, where the population is predomi-

nantly Hindu, Jokowi secured over 90 percent of the vote. Yet in a country that is
predominantly Muslim, non-Muslims can be no foundation for outright victory.
It was instead Jokowi’s sharply expanded support among largely rural Muslim
voters in Central and East Java that carried the day. The main difference between
the 2019 and 2014 votes was that Jokowi gained overwhelming support from
followers of the NU, especially in Central and East Java, after basically splitting
the NU vote with Prabowo in the 2014 campaign.
It is not hard to perceive how legacies of egalitarian nationalism helped make
this democratically fortuitous result possible. Any particular election result may
be highly contingent, and candidate-specific characteristics indeed help explain
why Indonesia defeated illiberalism at the ballot box in 2019 while India did not
(Slater and Tudor 2019). Yet if Indonesian nationalism did not make substan-
tial room for non-Muslims, it is hard to imagine how the pivotal PDIP-NU elec-
toral alliance could even have formed in the first place. The popular character of
Indonesian nationalism was also on full display in the country’s record turnout
in the 2019 vote: nearly 82 percent, spiking from an already impressive voter
turnout of just under 70 percent in 2014. It is thus clear that Jokowi was able to
tap into Indonesia’s deep reservoirs of egalitarian nationalism—a nationalism in
which Islam may play a major role, but the rakyat play the leading role—to save
both his own presidency and, arguably, Indonesia’s democracy from its avatar of
illiberalism, Prabowo.
Yet Jokowi himself has been increasingly tempted since his re-election to
deepen his own illiberal practices in turn. When urban protests erupted before
Jokowi’s second inauguration over bills passed by parliament that would gut
Indonesia’s independent anti-corruption commission, remove environmental
regulations on the mining industry, and restrict certain forms of political dissent,
the newly re-elected president unleashed police repression and internet controls
not seen since Indonesia democratized. This was not an electoral authoritarian
gaining a stranglehold on national power for himself and his ruling party by
restricting his electoral opponents. These were the actions of an illiberal dem-
ocrat, backed by the legitimacy of an emphatic double-electoral mandate, using
the full force of the state apparatus against society at large when it challenges
and speaks out against the political designs he shares with the broad swath of
Indonesia’s political elite. By subsequently inviting Prabowo himself to serve
as his defense minister, Jokowi signaled with abundant clarity that his second
term would be defined by the pursuit of economic development and political sta-
bility, not the continued defense of democracy. That job is once again being left to
Indonesia’s civil society, armed with an egalitarian nationalist spirit that provides
a strong defense, but by no means an unbreakable guardrail, against the ravages
of illiberal democracy.
Conclusion
Indonesia does not stand alone in the crosshairs of illiberalism. Anywhere

there are minorities to be scapegoated, dissidents to be intimidated, or lib-
eral institutions of constraint (e.g., parliaments, courts, anti- corruption
commissions) to be steamrolled, elected politicians will be tempted to gin up
popular support by doing so. The sad fact of the matter is that illiberal measures
are not only painfully easy for abusive leaders to impose—they are also often
painfully popular among a winning plurality of voters. The electoral side of de-
mocracy can mobilize the power of the many against democracy’s liberal side.
Our canonical theorists realized the possibility that democracy might under-
mine liberalism; sadly, more and more contemporary countries appear to be lit-
erally realizing this fate.
This is by no means a concern exclusive to Indonesia, where, more than in
many places, the electoral majority has rejected the most illiberal options on
offer. Illiberal forces in Indonesia are clearly on the rise, but not yet fully in the as-
cendant. In this respect, Indonesia looks very much like the main trendline in the
rest of the world. As fine-grained data from V-Dem show, electoral aspects of de-
mocracy have continued to improve around the world over the past decade. Yet
over the exact same period, the key liberal features of democracy—specifically
freedoms of expression and association—are quite vividly under global attack
(Ding and Slater 2021; Luhrmann and Lindberg 2019).
It is not only through extremist Islamic pressures to restrict Indonesians’ plu-
ralistic freedoms that illiberal democracy could tighten its grip. In a provocative
inversion of the conventional notion that religious pluralism bolsters democracy,
Fealy (2021) argues that Jokowi’s coercive turn against radical Islamic groups,
strongly backed by Jokowi’s allies in the military, marks the rise of “repressive
pluralism” in Indonesia. In this scenario, religious pluralism does not serve as a
wellspring of democracy but as a rationale for pluralism’s most avid defenders to
impose authoritarian measures, at least ostensibly to protect it.
It is not hard to see parallels for where this could lead. Containing Islamic
civil society has provided a ready pretext for military domination in cases like
contemporary Egypt and Kemalist-era Turkey. Even in Indonesia itself, one
might see the Suharto regime as a military-led dictatorship that was nearly
as concerned with bridling Islamists as communists. Indonesia does not
seem primed to fall into the pitfall of outright military takeover, for reasons
discussed above. Yet its military could easily become a leading player in an
illiberal democratic arrangement in which either political Islam is in the as-
cendancy and minorities are increasingly repressed, or political Islam is in the
elected government’s crosshairs and it is Indonesia’s religious majority that is
increasingly repressed.
This contrarian take on pluralism only reinforces this chapter’s argument that
illiberal democracy represents the most likely pitfall into which Indonesia might
fully fall. Indonesia is already worrisomely trending in an illiberal direction, but
it is still much further removed from an electoral authoritarian outcome, and a
much further cry from either military takeover or state failure, likely as both of
those might have seemed two decades ago. Any adequate theory of democratic
backsliding should distinguish among all these competing democratic risks.
Ironically, the same diversity that gives illiberal democrats a target-rich environ-
ment in which to hunt also makes it much harder for any one leader or party to
gain a stranglehold on electoral authoritarian power. If democracy keeps back-
sliding in Indonesia, we can expect illiberal democracy to tighten its clutches
first, particularly though the power aggrandizement of a domineering president.
Collective actors supporting either electoral authoritarianism or a full-blown
military takeover would then be waiting in the wings, sharpening their talons.
Notes
1. Indonesia has thus far fended off the challenge of exclusive, hierarchical nationalism
to its inclusive, egalitarian nationalist inheritance more successfully than India, for ex-
ample (Slater and Tudor 2019).
2. A fifth pitfall in Indonesia that I have discussed elsewhere at length (perhaps for some
readers, ad nauseum), what we might call elite collusion, is better considered a threat
to democratic quality than survival. If elite collusion is to lead to outright democratic
breakdown, it would need to do so through one of the other pitfalls discussed here.
Elite collusion at the levels Indonesia has experienced—through party cartelization
via promiscuous power sharing among ideologically diverse political parties (Slater
2018b)—is also considerably rarer than the other four pitfalls, making it less essential
for a conceptually and theoretically oriented intervention such as this.
3. East Timor’s secession gave birth to the new republic of Timor-Leste, discussed in
Nancy Bermeo’s contribution to this volume (Chapter 6).
4. This paragraph and its broader argument draw on my collaborative work with Maya
Tudor of Oxford University (Tudor and Slater 2016, 2020).
5. A brief comparison to democratic backsliding in the Philippines is informative.
Rodrigo Duterte is a consummate illiberal democrat, ruling in whatever manner he
sees fit. He also enjoys almost universal support among the political elite. But whether
he can parlay that support into assembling a coalition that undermines electoral fair-
ness and keeps either himself, his daughter, or some other close ally in Malacanang
Palace for more than just his single five-year term—thus crossing the boundary to elec-
toral authoritarianism—remains to be seen.
4
Africa’s Democratic Outliers
Success amid Challenges in Benin and South Africa
Rachel Beatty Riedl
Sub-Saharan Africa arguably poses some of the most challenging conditions for
democratic emergence and survival. According to our existing theories, this is
due to a number of structural conditions: low levels of economic development
and industrialization; regions of natural resource extraction; increasing levels of
inequality between political-economic elites and the masses; low levels of state
capacity related to colonial and neocolonial predation and post-independence
challenges to broadcast power; and multilinguistic, multiracial, multiethnic, and
multireligious populations. Yet three decades of democratic endurance in sig-
nificant, key cases suggest that the continent has much to teach us about how
democracies continue despite extremely challenging conditions. This chapter
explores two very different cases—South Africa and Benin—with highly di-
verse challenges and yet strikingly similar pathways to the establishment and
maintenance of democracy against all odds. In both countries, extreme polit-
ical, economic, and ideological crises made incumbent elites perceive the need
to concede their prior mandate for dominant control. Reactions against an intol-
erable past allowed new types of strategic alignments to be realized between eco-
nomic elites, security agents, ruling parties, and opposition elites. Transition-era
bargains for inclusive access to the state, security, political representation and
economic dominance resulted in a concession for democracy as a system of po-
litical inclusion and bounded uncertainty. In these cases, opposition elites real-
ized significant benefits to stability and order that could be had with including
the old regime in the new ruling coalition. Despite the challenges of a racially
or ethnically cleaved society in South Africa and Benin, the democratic bargain
was maintained through self-interested and embattled incumbents staring down
the precipice and opting for institutional compromise given the infeasibility of
maintaining the status quo and opposition elites’ preferences for maintaining ec-
onomic and social order that they could inherit when allowing the incumbent
some enduring role.
Democratic survival was sustained for as long as self-interested eco-
nomic and political elites saw the cost of institutional inclusion and bounded
Africa’s Democratic Outliers 95
political turnover as lower than the cost of potential insecurity (economic,

political, and personal) and zero-sum exclusion from the elite sphere of in-
fluence. Incumbent presidents of all ideological persuasions have been held
back from executive aggrandizement and personal impunity by the collec-
tion of counter-elites enforcing inclusive, informal, power-sharing models
of heterogeneous elite access to power. In South Africa, inclusive, informal
power-sharing has meant the gradual reformulation of the African National
Congress (ANC) leadership and accommodating internal factions; a growing
and diverse political opposition; and maintaining the buy-in of the apartheid-
era economic and political elite in the regime system. In Benin, informal
power-sharing has been achieved through the fluidity of the party system
with inclusive bandwagoning around independent candidates elected presi-
dent, and it has been funded by a state-aligned economic elite. When that bal-
ance of power became too concentrated—when economic and political power
rested in one central figure in Benin, President Talon—the counter-elite were
no longer able to contain the executive impulses to ravage democratic free-
doms and political opposition.
In sum, the tenable democratic compromise survives for as long as the
counter-elite are sufficiently diverse to raise the costs of democratic defection,
stemming from opposition political elites, economic elites, security forces,
workers, unions, media, and civil society leaders. Balancing against incumbent
overreach does not occur solely among opposition political parties; rather, op-
position parties’ ability to compete is a direct result of the commitment to de-
mocracy as the tenable compromise from all sides, in the absence of strong
institutional protections. Institutional design at the moment of transition helped
establish a framework for the endurance of such a tenable compromise. But
institutions alone cannot explain three decades of democratic survival, espe-
cially when constitutional change has occurred so rapidly and extensively in the
region, undermining protections such as term limits and civic liberties. The in-
stitutional framework follows the democratic bargain, and helps to maintain the
equilibrium for as long as the evolving political and economic distribution of
power can be managed within it.
The Puzzle of Democracies in Hard Places in Africa
African democracies present a real puzzle to most theories of democratization

and endurance. According to our existing theories of democratic origin and en-
durance, they are exceptionally unlikely to endure due to a number of structural
conditions: low levels of economic development and industrialization; pockets
of natural resource extraction; low levels of state capacity related to colonial
predation and postcolonial challenges in nation-building; and multilinguistic,

multiracial, multiethnic, and multireligious populations.
Yet the third wave of democratization swelled across Africa in the early
1990s, producing a set of political transformations that moved the vast ma-
jority of countries from military and single-party regimes to multiparty
regimes, with varying degrees of associated liberalization. This vast political
shift to multipartism was due to a combination of two interrelated factors
(Englebert 2002; Herbst 2001; Riedl 2014; Young 2012). First, the accumu-
lating consequences of the politics of economic crisis (van de Walle 2001) and
long-term economic stagnation of the 1970s and 1980s created tensions within
the ruling party in some cases, the disenchantment of civil servants and police
and military officers who were not getting paid, and the growing frustration
of the domestic population over the lack of development and public services.
This contributed to real pressures for reform from within, and the potential for
destabilizing protests, particularly in the capital cities (Bates 1984). Second, the
end of the Cold War, the defeat of the Soviet Union, and the associated chan-
ging priorities in international aid created external pressures for economic and
political reform to maintain foreign assistance and promoted the model of cap-
italist multipartism for contemporary political organization. The ascendancy
of the “West” and the neoliberal order meant that ruling parties or opposition
elites that had previously identified as Marxist-Leninist or received financial
support from the Soviet Union were particularly at pains to internally restruc-
ture and make new external alliances. The collapse of the USSR had profound
reverberations within African ruling parties.1 The geopolitical changes cata-
lyzed experimentation with alternative models for maintaining power, such
as new forms of competitive authoritarianism (Levitsky and Way 2010). Thus,
the move to multipartism was part and parcel of the shift from Cold War–era
single-party rule toward attempted new models of hegemonic control or var-
ious types of reform. This world-historical moment created a stimulus for
political change, a move to multipartism, and the potential for a democratic
transition. But it did not in any way guarantee enduring democracy (Bratton
and van de Walle 1997; Levitsky and Way 2010; Lynch and Crawford 2011;
Young 2012).
Given that the institutional changes to implement multiparty elections across
the continent had a variety of outcomes (Figure 4.1), there are empirical and
theoretical gains in understanding where and when democracy was maintained
against the odds. Two particular pathways led to democratic endurance. The first,
authoritarian-led democratization, was possible where incumbent ruling parties
had sufficient mobilizational, coercive, and/or material strength to manage
the transition on their own terms, shape the institutional landscape according
to their preferences, learn to lose, and reap the benefits of institutionalized
Pathways
Authoritarian-Led Democratization
(Ghana, Senegal)
1
Authoritarian Incumbent Concedes Defeat:

2 Opposition Allows Inclusion in Next Round
(South Africa, Benin)
Crisis/Status Quo Dominant
Control Unsustainable 3 Authoritarian Incumbent Party Employs Violence;
New Forms of Autocratic Regime Established
(Burundi, Zimbabwe)
4
Opposition Forces Employ Violence, Overthrow
Old Autocrats, and Take Power; New Forms of
Autocratic Regime Established (Côte d’Ivoire)
Figure 4.1. Regime transition pathways
competition and alternation (Riedl 2014; Riedl et al. 2020). In Africa, Ghana and
Senegal exemplify this pathway.
This chapter focuses on an alternative pathway for democratic endurance,
when the authoritarian incumbent must exit power at the transition and the in-
coming leaders pursue a strategic institutional inclusivity. In this pathway, the
authoritarian incumbent lacks the capacity to remain at the helm in the new
system, yet the rising opposition constructs a system that purposefully includes
remnants of the old order and guarantees continued access to political and ma-
terial power resources (Riedl 2018). This is the case for Benin and South Africa,
wherein groundbreaking and previously unthinkable transitions ousted the old
regime and empowered a set of transition leaders, who constructed a new insti-
tutional system and yet carved out inclusions based on a perceived need for con-
sensus rule to carry the country forward.
These transitions appear more revolutionary and more precarious, be-
cause the old order is dismantled and a new process of negotiation and inclu-
sion takes place. The transformation requires uncertainty about the nature of
the new compromises. New actors construct the democratic rules of the game.
Yet, in those countries that successfully maintain democracy, the new participa-
tory coalition allows the ousted incumbents to maintain a role, or to repurpose
themselves in new form that allows collusive power-sharing. The high political
uncertainty surrounding crisis-wracked transitions to democracy encourages
party elites to cope by sharing power and forging institutions to guarantee state
access across the country’s most salient political cleavages (Slater and Simmons
2013). While not in control of the transition, outgoing authoritarian incumbents
find safeguards for access and representation in the new system’s institutional
crafting. The rising opposition—the transition leaders—realize the benefits of
including the authoritarian incumbents both to help stabilize the new regime
economically and to avoid the historical shadow of chaos from deep ethnic and
racial divides. Paradoxically, in these two cases, the omnipresent fear of insta-
bility from racial or ethnic divides helps to forge a new inclusive coalition.
Driven to make power-sharing institutions to accommodate the old and
new elites, and sustain them for decades, these were contingent choices, but
historically structured. “While not determining which one of a limited set of
alternatives political actors may choose, . . . ‘confining conditions’ . . . restrict
or in some cases enhance the choices available to them . . . and even predispose
them to choose a specific option” (Karl 1990, 7). The deeply historical cleavages
in South Africa and Benin are structural constraints that shape both the authori-
tarian/apartheid period and the way in which democracy could be achieved and
maintained through compromise with old and new political elites. The repudi-
ation of the past regime created both a commitment to a new system as well as
a deconcentrated, diversified set of political and economic actors to be included
in its aggregation. Importantly, no single ascendant group was sufficiently pow-
erful to perceive its interests in dominating the political field—a commitment to
democratic endurance was born as a solution to managing conflict between an
outgoing but economically resourced elite and an incoming and internally fac-
tious elite.
Certainly, deep racial or ethnic divides do not generally augur for inclu-
sive coalitions. In pathways 3 and 4, authoritarian incumbents or opposition
members employ violence and limit participation, political rights, and civil lib-
erties, and new forms of autocracy are established (Adejumobi 2000). But given
these varied outcomes, and that democracy was never prone to begin or con-
tinue over time under these challenging conditions, the grave risk and historical
weight of racial/ethnic instability in Benin and South Africa helped drive rising
opposition to realize the strategic value of including the former authoritarian
incumbents in the new participatory, representative, and consensus-based po-
litical order.
South Africa and Benin established enduring democracy due to the interac-
tion of three contextual factors that catalyzed and maintained the inclusive coa-
lition. First, it was precisely the extreme political, economic, and ideological crisis
that made elites willing to concede their claims for hegemonic control and use their
incumbency to bargain for inclusion and benefits in the next system. These regimes
were on the brink of revolutionary overthrow, civil war, and/or state collapse, yet
they had not crossed the threshold. They retained a structural position to nego-
tiate, to concede with benefits. Incumbents became acutely aware that they and
the system that upheld them could not endure as it was. Instead, they settled for
a way out of the impasse and a system of electoral competition with formal and
informal guarantees that would protect their economic and physical security and
their ability to continue to be elite in the future, albeit without the possibility of
dominance.
Second, the historical experience of instability, violence, and political chaos
associated with racial (in South Africa) and ethnic (in Benin) divisions cre-
ated incentives for the rising opposition to prioritize security, governmentality
(Bayart 1989), and inclusive institutions in order to maintain a state that they
could rule over. This bargain for security at the height of a looming crisis in-
cluded broad power-sharing mechanisms, and citizens, political elites, economic
elites, and previously divided security forces all had a stake in supporting the
new system.
Third, significant redistribution was off the political agenda, and this agreement
among outgoing and incoming elites allowed the inclusive power-sharing pact to
endure. This argument at first glance seems to support the idea that democratic
transitions are most likely to occur when the costs of repression become greater
than the costs of inclusion, because authoritarian incumbents cannot maintain
themselves at the helm (Acemoglu and Robinson 2006; Boix 2003).2 However,
as with many democratic bargains, the demands for political reform and repre-
sentation did not entail a redistribution of economic power, and certainly not to
the masses. The democratic bargain included an enlarged political and economic
elite class; increased macro security, state functionality, and reduced external
and internal barriers to accumulation; and instituted elections as a mechanism
for expressing, channeling, and ultimately limiting citizen discontent. In these
cases, democracy is an institutional guarantee of the power-sharing bargain
among elite representatives of social groups defined by ethnicity or race, not a
class-based redistribution settlement.
Democratic Survival
So how did democracies survive against the odds? For thirty years in South
Africa and Benin, the compromise of inclusion among contesting elites—crafted
under duress, but institutionalized as durable power-sharing and rule through
consensus—facilitated democratic endurance. In South Africa, consensus rule
meant a commitment to maintaining white capital, and representation through
national Proportional Representation and elements of federal rule through pro-
vincial autonomy. In Benin, consensus rule meant low barriers to entry for po-
litical parties and candidates so that the former ruling-party members could
reconstitute in new ways, as well as constitutional and legislative protections
to require full agreement of all representatives before further changes could be
made to the system.
The economic, political, and social challenges they faced made them some of
the most unlikely to achieve stable democracy at the moment of multiparty re-
form, and yet the role of incumbent elites conceding to partial inclusion and in-
surgent opposition’s strategic power-sharing facilitated meaningful and lasting
democratic rule. The bounded uncertainty and rotation of elites—with little ac-
tual exclusion and manageable turnover of the top post—was deemed to be a
lower cost than the potential risks of physical insecurity, economic disruptions
due to strikes and protests, and full political exile. The all-or-nothing, zero-
sum stakes of the authoritarian era ceded to an expanded circle of all ideolog-
ical persuasions. While no one single economic or political opponent would be
strong enough to hold executive aggrandizement and presidential impunity at
bay, the coalitional weight of heterogeneous economic and political elites in both
countries was sufficient to serve as a counterforce to successive attempts by pres-
idents in Benin (Mathieu Kérékou, Boni Yayi) and South Africa (Jacob Zuma) to
stay in power and/or avoid accountability through increasing dominance.
Democratic survival in these cases demonstrates the interaction be-
tween agentic and institutional approaches. Certainly lacking the structural
foundations for democracy, South African and Beninese heterogeneous elites
(economic and political, outgoing and incoming) each perceived a strategic op-
portunity to maintain a democratic regime as a limited informal power-sharing
agreement. In doing so, they limited threats from below for greater redistribu-
tion and state services, suppressing social movement pressures for economic and
public service dividends by providing elite inclusion across the diverse spectrum
of representation provided by democracy. By maintaining representatives of the
old apartheid regime in the system in South Africa, and politicians associated
with the failed authoritarian single party in Benin, both countries avoided civic
conflict and limited physical insecurity to elite rulers, while also limiting poten-
tial for protest and economic disruption through the inclusive channels of demo-
cratic representation. This system maintained various degrees of political power
to the elite class; private property was maintained in South Africa, and lucrative
deals were made to control the port authority for economic elites in Benin. To
create further incentives to maintain this strategic equilibrium, and recognizing
its fragility, elites crafted institutions to help lock in mechanisms of inclusive
power-sharing.
Benin and South Africa today each face distinct challenges in democratic
endurance. In South Africa, the challenge to democracy stems from a rise in
radicalism, a willingness to consider non-democratic forms of rule linked di-
rectly to a legitimacy gap, and growing perceptions of corruption (de Jager and
Steenekamp 2019). These democratic deficits are represented by a slide in the V-
Dem Liberal Democracy Index from its height in 2007 at 0.68 to a score of 0.57 in
2019, but they are not severe enough to question the democratic compromise at
the elite level (Coppedge et al. 2020b). In Benin, the breakdown of the inclusive
power-sharing pact in 2019 stems from the overconcentration of economic and
political power in a single player—the current president, Patrice Talon. Talon’s
calculation to further centralize all resources in his “empire” and dramatically
limit opposition reflects a new calculation—that inclusive power-sharing and
the grand coalitions of the past are no longer necessary or preferable to main-
tain order, stability, and governmentality in Benin. His dangerous politics of ex-
cluding the opposition have led to dramatic democratic backsliding (Freedom
House [2020a] estimates a 13-point slide in Benin’s 2019 score compared with
2018; V-Dem Liberal Democracy Index shows a decline from a high in 2017 at
0.63 to 0.46 in 2019 [Coppedge et al. 2020b]). But Talon’s calculations may prove
the previous democratic transition strategy correct in the end: without inclusive
power-sharing, Benin teeters on the edge of political violence and a return to its
ethno-regional paralysis of the post-independence years.
Case Selection: Benin and South Africa as Africa’s

Democratic Outliers
This analysis focuses on two deeply divergent cases with shared general
challenges: Benin and South Africa. At one level, they are both “hard places” for
democracy to take root, in ways that resemble many countries across Africa and
globally, as described in this volume. The extreme challenges include a highly
diverse, multiethnic citizenry, challenges of state capacity in the periphery, and
relative levels of underdevelopment of a large section of the population. Colonial
legacies of arbitrary borders created states with little national cohesion and then
were administered by divide-and-rule tactics to further separate the population
(Lang 2009) and extraction to benefit the political class (Mamdani 2018). Power
was highly centralized in the executive, with few checks (van de Walle 2003). But
Benin and South Africa each confronted unique challenges to establishing de-
mocracy as well.
In Benin, the post-independence history of tripartite ethno-regional divisions
and deadlock contributed to extreme political instability, with six coups and as
many government breakdowns and new constitutions in the first twelve years
(Decalo 2018; Magnusson 2001). The underlying ethno-regional fragmentation
and an equally divisive economic and social disparity between north and south
created a perpetual crisis. Benin was temporarily stabilized with the final mili-
tary takeover by Mathieu Kérékou in 1972, and the transformation to a nominal
Marxist-Leninist single-party regime, buttressed by the military. The coercive
power of the state was put to use as a threat to society and individuals through
the suppression of political and civic activities, arbitrary arrest and torture,
suppression of local religious and cultural practices, and generalized assaults on

property rights (Challenor 1970; Ronen 1975). This combination of fragmented
society and coercive military/personalist/single-party rule did not provide an in-
stitutional foundation for successful democracy (Banégas 2003, 2014).
Equally critical, Benin’s economic conditions did not provide fertile ground for
a flourishing democracy. In 1990, the country ranked near the bottom of global
levels of development, at a level of $1,230 GNI per capita, PPP (current inter-
national $, World Development Indicators, n.d.) and was facing massive public
debt. In addition, it was reeling from the latest oil-price decline shock in the
1980s, in which neighboring Nigeria sharply contracted its demand for informal
trade, the main source of sustenance for much of Benin’s population (Magnusson
and Clark 2005), resulting in significant economic decline (Figure 4.2).
The state economy was a classic rentier state, in which material benefits are
closely related with access to political power, and maintaining power is there-
fore necessary for one’s economic condition (Bierschenk 2009). “By the late
1980s the government was bankrupt, the banking system was looted, paychecks
were months in arrears, and the civil service was on strike. The state virtually
collapsed. By any historical or economic standard, Benin was a poor prospect for
democracy” (Magnusson 2001, 218). Mired in weakness and instability, with the
specter of past ethno-regional cleavages, Benin’s challenges were daunting.
In contrast, in South Africa, the headwinds to establish a peaceful democ-
racy were relatively different, and largely distinct to the racialized legacies of the
10,000
9,000
8,000
GDP per capita (2011 US$)
7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994
Benin South Africa
Figure 4.2. Pre-transition economic decline in South Africa and Benin, 1980–1994
Source: Coppedge et al. 2020b
apartheid system and competitive oligarchy. Whereas Benin lacked an institu-

tional basis of multiparty competition, but had mobilized—even demanded—
full participation under the single-party regime, South Africa had established
vibrant multiparty democratic competition and institutional mechanisms for
the transfer of power within an extremely limited portion of the population
(Heribert and Moodley 1993, 20; Southall 2001, 22; Thiel and Mattes 1998, 101).
This difference highlights Dahl’s (1973) paradigmatic differentiation between
pathways of limited participation with high competition and those which had
high levels of participation but limited competition.
Rather than democracy, most external analysts and many internal players
were preparing for the prospects of violent revolution in the late 1980s, and “very
few predicted the possibility of a soft landing for apartheid” (Herbst 2003, 107).
Even after the founding post-apartheid elections in 1994, many thought democ-
racy was not possible, arguing that “South Africa lacks the social requisites as-
sociated with the consolidation of liberal democracies: broad-based economic
development and national homogeneity” (Giliomee 1995, 84). Others recog-
nized that despite the Truth and Reconciliation Commission’s attempt to deal
with apartheid-era justice claims, the police and justice system’s systemic racism
was a barrier to overcome in the transition to democracy (Shaw 1997). Many
feared extreme polarization following the negotiated settlement, as the centrist
compromise would be unable to contain the stratification of the goals of black
and white South Africans (O’Flaherty 1992).
South Africa’s economy was established and relatively strong on the conti-
nent, with large corporations diversified across various sectors (Nathan 2004, 5),
providing overall significantly higher levels of development at $6,180 GNI per
capita, PPP (current international $) in 1990 (World Development Indicators,
n.d.). Far from a “failed state,” South Africa was effectively integrated into the
global economy and had a manageable debt (Habib and Padayachee 2000, 247).
Yet the country was also facing significant economic decline throughout the
1980s (Figure 4.2). The economic implosion was due, in large part, to rising in-
ternational sanctions against the apartheid regime, levied on South African
companies, and decreasing investor confidence in the face of rising domestic po-
litical turbulence, which made the status quo unsustainable (Giliomee 1995, 89;
van Wyk 2005, 53).3 And the extreme levels of inequality meant that the vast
majority of the population was both economically impoverished and politically
disenfranchised. South Africa’s economy was completely dependent upon white
capital. The Gini Index value in 1993, prior to the transition, was 59.3, one of the
highest in the world (World Development Indicators, n.d.). This concentrated
distribution created major challenges for democratization, because the white mi-
nority feared total personal insecurity and economic appropriation with the loss
of political power. And many within the black majority recognized the need for
economic stability but also faced strong demands for reorienting the economy
to address the structural apartheid of labor and capital control. Rising domestic
protests created disruption in the supply of labor and productivity at the macro
level and increased costs of state monopolization of force. This economic insta-
bility was both a challenge and a catalyst for the elite compromise necessary for
democratic reform. The vastly unequal distribution of wealth suggested the huge
risks the ruling apartheid regime faced from democratization, yet the increasing
sanctions and domestic pressure made the existing political and economic situ-
ation untenable.
Crucially, despite the vast differences in economic distribution and social
cleavages, in both countries it was not possible for the ruling party to main-
tain hegemonic political control. The fiscal pressures created splits within the
ruling classes. In South Africa, the white ruling class had important differences
between the British and Afrikaner populations, with the Afrikaners generally
supporting a “Nationalist Unity” agenda for apartheid, although this cohesion
was precarious even within the Afrikaner population (Lieberman 2003; Welsh
2009). Whites who were more oriented toward external markets, and British
with foreign passports which guaranteed them an exit option, moved toward re-
gime transformation prior to the Afrikaners. Benin’s fiscal crisis meant that civil
servants, police, and other elements of the security apparatus were not getting
paid, and these agents of the state started to turn against the ruling party. These
economic pressures on the elite were coupled with massive discontent and rising
protests from the population for reform. Both South Africa and Benin were
facing political, economic, and ideological crises, and stakeholders on all sides
sought reforms to allow macro political stability, a new ideological foundation
for rule, and the foundations to regenerate economic opportunity. These were
not regimes that could rely on state control of natural resources; instead, there
were demands from all sides for a functional state and political reforms to facili-
tate future economic stability and growth.
Today, both South Africa and Benin represent incredibly divergent modes of
democratic durability. Benin’s democracy has produced four executive power
transitions of individual and political party, demonstrating high levels of vol-
atility, competitiveness, and participation. During the past three decades fol-
lowing the transition, Benin ranks solidly in V-Dem’s Liberal Democracy Index,
given the highly competitive political landscape and political rights and civil
liberties accorded through 2018 (Figure 4.3). The institutional separation of
powers, and in particular the autonomy of the courts, the role of civil society in
mobilizing against democratic threats, and the dizzyingly fluid playing field of
candidates indicate high levels of contestation and participation on the part of
the elites and citizens alike. According to 2016–18 Afrobarometer data (Round
7), 80 percent of the population rejects single-party rule, and 72 percent say that
0.8
0.7
V-Dem Liberal Democracy Index
0.6
0.5
0.4
0.3
0.2
0.1
0
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
South Africa Benin
Figure 4.3. V-Dem liberal democracy index rankings in South Africa and Benin,
1960–2019
Source: Coppedge et al. 2020b
democracy is preferable to any other form of government (only 13 percent think

“it doesn’t matter what kind of government we have”). Of course, the fluidity of
the democratic playing field creates challenges for accountability, but citizens
had the ability to vote for an array of candidates and sanction those that were not
performing, creating extremely high levels of turnover in the National Assembly
(at times, over 70 percent).
Benin’s recent democratic backsliding in 2019–20 marks an attempt to end
the consensus rule and inclusive coalition that sustained democracy over the
prior period. As economic and political power became overly concentrated in
the executive, the current president has endeavored to cut out opposition (Preuss
2019). Because President Talon himself was a major funder of a diversity of po-
litical parties in prior governments, and mobilized the civil society movements
against earlier constitutional changes and potential term limit challenges, his
strategic role in maintaining a fluid and rotating political elite field was critical.
When, however, his economic interests (and particularly his control over the lu-
crative port) were threatened by his erstwhile ally, President Boni Yayi, Talon
recalculated that his economic dominance and security required direct political
control, rather than indirect proxies. Once Talon assumed the presidency, the
combined force of the political opposition and economic competitors was largely
insufficient to contain his crackdowns on civil liberties and political rights. The
diffusion of economic and political resources helped Benin’s fragile democratic
equilibrium maintain against all odds, but with Talon’s ascension to executive
control, there are few checks on his dominance. Whether the threat of political
instability will force a return to competitive inclusion is an open question for

the future. The question itself is an indicator of the role that Benin’s historic
challenges have played in undergirding the democratic compromise for inclu-
sion and stability.
In South Africa, V-Dem ranks the country high in the Liberal Democracy
Index (0.57) (Figure 4.3), and Afrobarometer survey data show that in 2016–
18, 62 percent of the country disapproved of single-party rule, and 54 percent
preferred democracy to any other system (whereas 25 percent agreed with the
statement that “it doesn’t matter what kind of government we have”). The sup-
port for democracy is striking in South Africa because it reflects both a liberation
struggle for equality and rights and an attempt to make political progress on the
legacies of apartheid. The ANC has ruled as a dominant party and the party of
liberation since the first post-apartheid elections. The party has also seen sig-
nificant internal rotation with five separate presidents elected. The strength and
social rootedness of the ANC have provided stability to the democracy, allowing
institutional bargains of inclusiveness to be worked out in practice. But the lack
of party alternation in the executive has also limited accountability and opposi-
tion voice. The strength of the ANC undermines the constitutional separation of
powers and the accountability of the executive to the electorate (Seekings 2009).
Yet the formal procedures of representative democracy remain deeply significant
for a range of social group interests to be expressed and for transformation to
occur over time.
Forging and Maintaining Democracy
The question both Benin and South Africa faced as they began to embark on
reforms was the degree to which the former autocratic rulers (and the identity
groups they represented) could concede power while maintaining institutional
guarantees for representation, economic security, and continued political veto
and agenda-setting powers. Unlike many expectations and fears of majority rule
and massive appropriation, the ruling elite found that moving to democracy
did not entail a full concession of power. Rather, democracy entailed new stra-
tegic bargains over including the former elite and mapping out rules of the game
that would protect certain perquisites. The road to increasing participation in
South Africa and contestation in Benin was paved with negotiations around the
preservation of formal and informal access to the political center. Institutional
foundations (federalism in South Africa) and informal practices (of the vast ma-
jority political parties bandwagoning into the presidential coalition in Benin)
were crafted to provide inclusion for the former political elites and the groups
they represented along with the ethnically diverse newcomers.
Crisis, Cleavage, and (No) Redistribution as Context
Democracy was never a foregone conclusion, much less its endurance. In both
countries, civil war loomed large and political violence was omnipresent. Society
was highly mobilized for reform through domestic riots. Political inclusion and
competition—democracy—was the first core priority. For some, democracy was
an end in itself. There were absolutely ideological believers who have been com-
mitted to maintaining a democratic polity: Nelson Mandela in South Africa and
Nicéphore Soglo in Benin stand out as identifying democratic principles as the
necessary solution for their respective countries’ peace and harmony. For others,
democracy was a means to an acceptable end of economic and physical security
and partial political representation rather than full. Particularly for outgoing ec-
onomic, political, and security elites, their changed positionality and strategic
reassessment through crisis and repudiation of the past was what allowed for a
democratic bargain to be struck and maintained.
In South Africa, economic empowerment was tightly intertwined with polit-
ical rights and the end of the apartheid system. Claims for redistribution and
reshaping a structurally unequal economic system were central, but they were
also balanced by a pragmatic agenda to stabilize white-owned capital and harness
it to the benefit of all citizens (Emery 2006). Among the apartheid white ruling
class, the critical shift was from hegemonic political control to institutional guar-
antees for political overrepresentation and economic ownership. The new con-
stitution adopted a form of federal governance that provides protection for the
economic elite from maximal redistributive taxation (Inman and Rubinfeld
2012a). Among the black majority of the national liberation movement, the key
shift was from revolution to democratic participation, legal equality, and rec-
onciliation. Negotiations struck a crucial balance between advocating for their
position and making concessions. In doing so, they alleviated white fears (e.g.,
the ANC specifically ensured whites of their jobs, pensions, and farmland) while
also maintaining black support for the transition (Welsh 1994, 222–23). While
not addressing inequality, the system improved the economic welfare of both the
white minority and the black majority, contributing to the enduring support for
democracy.
In Benin, economic crisis meant that the ruling elites also gave up their claims
to hegemonic control. In 1990, mass mobilizations and protests forced the ruling
party to accept pressures for reform. They held a National Conference as a way
to negotiate the crisis, which was meant to give voice to all sectors of society,
to vent their frustrations. The goal was to increase temporary participation and
voice among stakeholders throughout the country, and re-emerge in control. The
opposition declared their own sovereignty, but they too made limited demands.
They allowed the sitting president to maintain his position as head of state. They
prioritized political foundations of constitutionalism, democratic contestation,

and social participation over any questions of redistribution. Group claims were
not made on the basis of class but on the basis of regional, ethnic, or associational
representation.
In sum, political- social-
economic- ideological crises provided a foun-
dation for far-reaching reforms and for the crafting of democracy. The non-
democratic ruling elite needed to be in a structurally untenable position; the
pro-regime coalition was cracking from within under pressure. But these
conditions were far from sufficient. In many other cases across the continent,
these conditions proved ripe for civil conflict and new forms of authoritarian
rule. In both South Africa and Benin, recent history and the threat of political
instability cleaved along ethno-regional or racial lines upped the ante of costs
and benefits of competition over power. Given the high probability of disorder,
the insurgent opposition had a strategic use for maintaining and including the
declining ruling elite in some form. Whether for economic stability and poten-
tial future growth in South Africa or for sociopolitical stability in Benin, the
old guard was maintained in inclusive institutional guarantees. This avoided
destructive conflict and allowed democratic provisions of electoral representa-
tion as an institutional fix. Political channels for participation and representa-
tion (and therefore economic protections against democratic theory of drastic
redistribution through expropriation) were critical to democracy’s founding
and endurance.
South African Negotiated Compromise
Apartheid’s long history in South Africa set the stage for a deep institutional
and social cleavage that had to be addressed in order to achieve durable de-
mocracy.4 The increasing violence and social, political, and economic disorder
throughout the 1970s and 1980s created high animosity and distrust between
the governing elite and the mobilized black majority, but it also laid bare the
stark reality that a solution could not be zero-sum (Nathan 2004, 3–4). For South
Africa to avoid civil war and implosion, incumbents and opposition increasingly
realized, however painfully and against their ideological preferences, that a ne-
gotiation of some kind would have to take place. By 1979 the South African gov-
ernment was producing reports about labor laws and worker segregation that
ultimately viewed apartheid as unworkable, due to the increasing migration of
blacks into “core areas” (white South Africa) (Wiehahn Commission 1979; cited
in Herbst 2003, 100).5 Throughout the 1980s, white-led reform was carried out
largely by Afrikaner nationalists in power, with the objective to avoid the “crea-
tion of a multiracial government, but [ensure] the preservation of effective White
control” (CIA memorandum, January 1981; quoted in Herbst 2003, 101).6 There
was widespread recognition that Blacks, Coloreds, and Asians recognized the
intention of these reforms, further fueling tactics of protest and township riots
between 1984 and 1986, and labor union strikes to exert pressures on the gov-
ernment. An emerging opposition within the Afrikaner National Party in favor
of more substantive changes lacked influence in the party itself or in Parliament,
but began to create divides within the party as well as the broader community.
President Botha sought to maintain apartheid and was ideologically against an
integrated government. Yet structural forces pushed negotiations along without
him; economic discussions began in the business sector between white business
leaders and the ANC, along with academics and other professionals. Ahead of
the political leadership, they sought to refine business strategy, regain a qui-
escent labor force, and increase cooperation internally and externally (Marais
and Davies 2015).7 Estimates of the state’s capacity to exert sufficient coercive
force meant that they were capable of avoiding democracy: “Pretoria’s economic,
political, and military resources are sufficient to avert the collapse of the white
regime” (CIA Special National Intelligence Estimate, August 1985; quoted in
Herbst 2003, 102).8 So while the costs of repression were mounting internally,
the regime had capacity to maintain apartheid through coercive force, yet many
began to see the imperative of change in one way or another. Botha’s stroke in
1989 led to change in the leadership of the National Party (NP), and F. W. de
Klerk’s rise signaled less of an ideological shift than a recognition that, given the
growing unrest in South Africa, “a clear departure from the policies of the past—
one way or another—was needed” (O’Brien 2017, 627).
De Klerk is emblematic of the structural position of the white ruling class—
embattled and facing the need to reform, but with a wealth of resources at his dis-
posal to negotiate in the interests of his constituency’s enduring power. De Klerk
rose to the leadership of the National Party and then state presidency in 1989,
known simultaneously as a strategic reformer and for his conservative attitude.
His past ministry positions suggested that he “did not decide to negotiate an end
to apartheid because of a lifelong commitment to liberty and equality. Rather, he
understood that economic decline and a changing international climate meant
that the situation he inherited was unsustainable. . . . [He] came to realize that
in such a context, his best chance of protecting the interests of his supporters
was to lead, rather than to resist, change” (Cheeseman 2015, 108). Yet there was
much to constrain him in imagining a democratic future in South Africa. He said
that “change would take place without endangering the ‘values and achievements
of communities and individuals of the past and our dreams and future plans’ ”
(Giliomee 1995, 94). De Klerk tried to play down the significance of changes in
order to “reassure his white constituency that the change underway was not all
that threatening” (Saunders 2009, 162).
An important structural change coincided with the transition from Botha to

De Klerk and contributed to the end of apartheid and the enduring democratic
compromise of South Africa: the dissolution of the USSR and the fall of com-
munism as a defining threat on the international stage. Without Soviet support
for the ANC, the Frontline States (a coalition of African states committed to
ending apartheid and white minority rule) exerted more pressure on the ANC
to negotiate instead of using violence (Rantete and Giliomee 1992, 527). The
fall of the USSR did not change the ANC’s ultimate goals or leadership, but it
did change their tactics away from violent confrontation and oriented them to-
ward compromise. De Klerk also admitted that the collapse of communism in
Eastern Europe opened up the possibility for substantive negotiations with the
ANC, as Botha, and to a lesser degree De Klerk, saw serious negotiations with
a communist movement to be unacceptable, as the South African government
sought to align itself with the Western world (Rantete and Giliomee 1992, 518–
19). When the ANC was receiving significant funding and militant support from
the USSR, was made up of an important segment of Communist Party members,
and was championing “drastic agrarian reform, widespread nationalization of
key conglomerates and ministries, and radical improvements in the conditions
and standards of living for the working people,” the West was hesitant to pledge
full support behind the ANC (Habib, Pillay, and Desai 1998, 103; see also Habib
1995). However, with the fall of communism, the ANC lost their main source
of support and was essentially forced to embrace neoliberal free-market prin-
ciples and shun more radical state interventions (van Wyk 2005, 150). South
Africa’s established business elite furthered negotiations in this wake, making
the costs of a radical interventionist economic program appear even higher and
more impractical (van Wyk 2005, 151). The increasing contact and communi-
cation between white business leaders, NP moderates, and ANC leadership be-
came crucial to building trust between the main negotiators. The Western and
Southern Africa regional orbit of multiparty systems and elections in the early
1990s further paved the road for an inclusive, negotiated transition, rather than a
radical agenda and zero-sum confrontation (Nathan 2004, 2).
The ANC’s experience as a virtual government in exile for multiple decades—
operating from neighboring Tanzania and Zambia—was also key to its ability to
envision an institutional bargain that would absorb the majority of white apart-
heid elites. This extended governance, experience over time, created a flexible
internal organization capable of integrating a variety of homeland-based polit-
ical parties and civic bodies established during the years of ANC exile, returning
soldiers, and varied types of liberation bureaucrats returning to play a role in the
transition and newly established democratic government.
The strength of domestic opposition to apartheid, and notably the ANC rev-
olutionary anti-apartheid movement, sought full participation and citizenship
rights as primary objectives of reform. The anti-apartheid movement was

highly mobilized, internationally connected, and ideologically oriented to end
white minority rule (Cheeseman 2015, 106). But the ANC itself was also in-
ternally divided, with different factions representing varied visions of a post-
apartheid world and the means to achieve it, and different views about the
value of liberal, multiracial democracy. The South African Communist Party
(SACP) had a considerable degree of influence, particularly within the mili-
tary wing of the ANC, and maintained its insurrectionary predisposition. The
Congress of South African Trade Unions (COSATU) also maintained a his-
toric alliance with the ANC and, through their strike capacity, furthered the
strategic and economic case for a democratic resolution. Breakaway factions
such as the Pan-Africanist Congress supported a more radical African racial
identity. Nelson Mandela was a key leader, strategically and symbolically. He
combined the urgency and necessity of liberation with the recognition of the
structural constraints that must be addressed to achieve it. Mandela and his
co-patriots in exile (Oliver Tambo, President of the ANC, Thabo Mbeki, Jacob
Zumba) began a negotiation that evolved in relation to the declining posi-
tion of the white minority’s perceived ability to benefit from the status quo.
ANC leaders pursuing this negotiation identified the structural inequalities
that had contributed to four decades of impasse, and ultimately accepted the
contradictions that overcoming the stalemate would entail in maintaining the
incumbent elite politically and economically, and committed their side to a
negotiated settlement (Welsh 2009). By the late 1980s the ANC leadership was
strategically and coherently planning for a post-apartheid liberal democracy,
and they presented this plan as a cautious move toward a negotiated settlement.
ANC leaders like Thabo Mbeki, the head of the ANC’s directorate of interna-
tional affairs in exile, and Mandela, from prison, began a series of meetings
with senior government officials and cabinet ministers to begin to craft an in-
stitutional blueprint for an inclusive settlement that could be acceptable to the
white ruling regime.
The ANC’s position, and their ability to negotiate for compromise, was shaped
by the structure of society. On one level, the pre-apartheid model of social con-
sensus seeking was crucial to inform a shared model of democracy and legitimate
the negotiation process that led to democratic elections, the negotiation strategy
of the ANC, and the settlement that emerged from it (Nash 2002, 253). Mandela’s
transformation of this consensus-seeking social order “legitimated the ANC’s
role as interpreter of the African consensus on the basis of the sacrifices of its
leaders” (ibid., 253). On another level, the apartheid organization of society was
made up of “sectors”—youth, women, business, labor, political parties, religious
and sporting bodies, and the like—each with a distinctive role to play. This idea
emerged from the organizational needs of the struggle against apartheid when
repressive conditions prevented them from mobilizing directly around political

demands. It was now used to “insulate the leadership of the liberation movement
from critical questioning” and thus allowed them to seek compromise and nego-
tiation (ibid., 253). Mandela explained to the Consultative Business Movement
with white economic elites, in May 1990, that “Both of us, you representing the
business world and we a political movement, must deliver. The critical questions
are whether we can in fact act together and whether it is possible for either one of
us to deliver, if we cannot and will not cooperate” (CBM 1997, 57–58). In calling
upon business, and in turn labor, youth, students, women, etc., to act within the
limits of a “national consensus,” the question of the basis of that consensus could
be removed from sight (CBM 1997, 58, 65, 66). “In effect, the ‘tribal elders’ of
South African capitalism were gathered together in a consensus that could only
be ‘democratic’ on the basis of capitalism” (Nash 2002, 253). White South African
business elites played a decisive role as midwives and sustainers of democracy—
just as they had been the staunchest allies of the oppressive apartheid regime—
because they sought to create a new political and economic climate to protect
and maximize the gains accrued in a highly uncompetitive economy (Makgetla
and Shapiro 2016).
This consensus was maintained during the transition’s negotiated settlement
process (the Convention for a Democratic South Africa and the subsequent
Multi-Party Negotiation Process) and throughout the continuing decades of
democratic endurance. In overcoming the possibility of revolution, the inclu-
sive model of democracy formed the ideological contradictions of the new South
Africa. At the outset of the institutional negotiations, preferences were extremely
varied: “the reformists among the regime camp, particularly the government. . .
favoured some form of power-sharing model to provide statutory guarantees for
the rights of minorities, while the hardliners in the regime camp . . . favoured
federalism and some form of self-determination respectively in order to protect
the interests of at least some cultural groups. Moderates in the anti-regime camp,
such as the ANC, favoured majority rule as the best model, but the anti-regime
radicals . . . favoured black domination. These were the first preferences of the
various actors,” but through negotiation these were moderated to compromise
over institutionalized power-sharing (Seo 2008, 452). Negotiations broke down
many times over the government’s preference for power-sharing and the ANC’s
preference for majoritarianism, and in these instances international mediation
played a role in overcoming the deadlock (ibid., 452). The threat from incumbent
hardlines, opposition radicals, and the further escalation of violence also pro-
vided structural incentives to find compromise, establish an interim government
of national unity, and institutionalize power-sharing. Ultimately, the institutional
foundations of guaranteed representation, land and wealth protections, and ju-
dicial independence provided both sides with more confidence that a negotiated
settlement could produce a lasting outcome and, therefore, represented a viable

path forward (Cheeseman 2015).
This commitment is institutionalized formally through a two-tier version of
proportional representation (PR) (Barkan 1995). Under the South African elec-
toral system, two hundred seats in the National Assembly are allocated on the
basis of the party vote nationwide, and two hundred are allocated on the basis
of the party vote in each of nine regions. This version of PR has resulted in each
party’s gaining a proportion of seats that is virtually identical to its percentage
of the total vote, with the largest parties guaranteed participation in the govern-
ment proportionate to their vote, and thus, proportionate to white representa-
tion in society when voting along racial lines.
The negotiated settlement also institutionalized power-sharing in other
formal ways: the parties agreed to a Government of National Unity (GNU)
for the first five years after elections, and the interim Constitution provided
for a series of power-sharing arrangements. It entailed that any political party
securing a minimum of 5 percent of the national vote would be guaranteed a
cabinet post, and any party which secured at least 20 percent of the national
vote was entitled to nominate a deputy president of the republic (Booysen and
Masterson 2009).
Similar power-sharing arrangements exist at the provincial level, and the
semiautonomy guarantees regional strongholds to the white minority, such as
the Western Cape. “The federal elements eventually built into the 1993 and 1996
Constitutions . . . and the state that emerged . . . should be seen as a negotiated
compromise, rather than the product of a single clear vision,” and dispersed
power among three spheres of government (national, provincial, and local) to
articulate certain principles of cooperative government (Steytler 2005). The pro-
vincial boundaries were explicitly negotiated to assure the white (NP) and black
(Inkatha Freedom Party representing the rural blacks of the historic Zulu na-
tion) political minorities control over public resources and policies in at least one
province each (Inman and Rubinfeld 2012b).
The federal system was critical to institutionalizing power-sharing and guar-
anteeing non-redistribution over time, at the heart of democratic endurance.
The form of federal governance established in the constitution provides “protec-
tion for the economic elite from maximal redistributive taxation. Appropriately
structured, federal governance creates a ‘hostage game’ in which the majority
central government controls the tax rate but elite run province(s) control the
provision of important redistributive services to a significant fraction of lower
income households” (Inman and Rubinfeld 2012b). The federal governance
structure therefore provided sufficient protection to the economic elite to sup-
port democracy, and it has contributed to overall gains in the economic well-
being and quality of life, including access to services, as well as the rise of a new
black middle class, further reinforcing the economic and political interdepend-
ence of the society (Munyeka 2014; Southall 2018; Wieczorek 2012). The basic
institutions of capital markets, private banking system, and contract law were
well established and maintained through this negotiated process, providing an
economic basis for neoliberal reforms to be rapidly and successfully enacted
(Bratton and Landsberg 1998).9
The consensus model requires the majority to sacrifice in the cause of building
and maintaining a democratic system, but it preserves the role of capital and
the power of incumbent elites in the process. It informs the “real relationships
of power behind the facade of formal democratic procedures. In its many in-
stitutional embodiments. . . it calls upon the oppressed majority, in particular,
to sacrifice in the cause of building a new society” (Nash 2002, 254). This con-
sensus preserves the role of the chief intact—from Mandela to Mbeki to Zuma
to Ramaphosa—and is based upon ties of solidarity in the common struggle of
apartheid for the black majority, and the elite consensus on power-sharing and
lack of redistribution.
The consensus model has also been embodied in the country’s approach to
justice, through the Truth and Reconciliation Commission’s non-retribution,
reconciliation paradigm. The Truth and Reconciliation Commission sought to
address both victims and perpetrators; justice would be restorative, not criminal,
to provide amnesty and reparations. This approach individualized the process,
reducing apartheid from a relationship between the state and the entire com-
munity to one between the state and individuals (Mamdani 2002). Similarly, the
establishment of a constitution and Constitutional Court that was legitimate
in the eyes of most South Africans was crucial to achieving successful demo-
cratic endurance. This legitimacy had to be based upon the freedoms equally ap-
plied across all members of society and the limitations it placed upon abuse of
authority—equally capable of restraining the incoming ANC as the outgoing NP
(Klug 2008).
As this model has continued over time, and through various presidencies,
some have argued that the continuation of the initial mandate to share repre-
sentative power undergirds the unity, stability, and, ultimately, democracy in
South Africa. For example, the Government of National Unity was designed to
provide ministerial posts to the new opposition (the NP and its allies); Mandela
and Mbeki continued the practice, despite being under no obligation to do so
(Nathan 2004, 4–5), in order to maintain the precious balance between capital
and state. The opposition NP eventually decided to disengage from this practice
in order to more effectively differentiate in electoral competition, but it took up
the practice of appointing a Shadow Cabinet to operate in parallel in order to
hold the national executive to account, oppose problematic legislation, and pro-
pose credible alternatives before Parliament.
The vibrancy and maintenance of South Africa’s democracy will likely not
be threatened by the racial cleavage that obstructed its origins. If anything, the
emerging threat to democratic endurance is from within the ANC, frustration
with their corruption, and their dominant hold on power. Of course, the ANC’s
dominance is fundamentally linked to its legitimacy as liberators of the country,
providing human rights and freedom through the end of apartheid (Booysen
2014). And their inability to address redistribution has given rise to the more
radical Economic Freedom Fighters, whose platform advocates for extensive
nationalization of industry. The opposition’s evolution into a more multiracial
party, the Democratic Alliance, suggests the evolution of a stable competitive
landscape. And the ANC’s ability to undertake internal reform, including leader-
ship change and accountability measures, suggests another positive indicator for
continued democratic endurance in South Africa. These elements continue to be
based on the institutionalized premise of non-retribution, reconciliation, power-
sharing, and representation that make up South Africa’s enduring democracy.
The stability that democracy provided was part of the glue that keeps it
maintained. Emerging from the precipice of economic and political upheaval,
South Africa’s inclusive bargain highlighted the gains to be shared by a new,
broader elite class and the societies that support them. Yet this bargain also
took off of the table a political revolution and socioeconomic restructuring.
Democracy’s bargain formally avoided massive redistribution. More radical
members of the trade unions, the ANC, and the Communist Party were frus-
trated by this settlement. But, in South Africa’s enduring democracy, they con-
tinue to contest and push the boundaries of that bargain. Indeed, the compromise
of South Africa’s democratic founding maintains in the shared interests of elites,
the interdependence of labor and capital, and of incumbents and opposition, to
prioritize stability, macroeconomic growth, and human rights and democratic
participation as the fruits of liberation.
Inclusive Coalitions in Benin
Benin’s ruling elites were not worried about expropriation in the way that those
in South Africa were. In Benin, as in most of sub-Saharan Africa in the late
1980s, private markets and avoiding tax capacity were not the main sources of
elite accumulation. Calls for a democratic transition were not about the transfer
of resources from an elite group to the masses. Rather, political elites and their
associates became economic elites by controlling access to state resources;10
maintaining access to the state was the core concern.
In Benin, ruling elites and citizens were worried about a more immediate
and generalized breakdown of the political and economic system as the country
faced a gripping crisis. And whereas South Africa’s compromise of medi-

ated demands gave whites overrepresentation in the political sphere through
formal institutions, in Benin the inclusive coalition was largely informal but was
buttressed by lower barriers to entry for new coalitions of recycled and ascendant
elites. The ruling party was dismantled, and the former ruling party elites dis-
persed into new and highly fluid constellations.
Much like in South Africa, the key players included the ruling party, forced
by its changing economic and political situation to grapple with reform, a
rising and diverse set of opposition elites (some who had been abroad during
the authoritarian period), and a broad array of civil society actors including
union representatives, religious groups, youth associations, market women,
and so on, who helped to agitate for change and demand inclusive reforms.
International actors were similarly but minimally involved in that they no
longer condoned the ruling party’s authoritarian hold on power and began to
restrict funds (in Benin through withheld foreign aid, rather than the sanctions
imposed on South Africa), and in doing so contributed to a transition toward
multipartism but did not require the ruling party to lose power or to create a
level playing field. As will be discussed comparatively in the following section,
the international influence at the time was one of several factors that contrib-
uted to the shifting perceptions of the ruling party’s diminished ability to sus-
tain itself and increased the party’s openness to consider negotiated reform,
as well as opposition’s sense of possibility to establish a new order during the
transition, but this democratic inclusion had to be sustained by domestic ac-
tors throughout the subsequent decades in order for democracy to endure
(Gazibo 2005, 2013).
As the economic, political, and ideological crisis of Benin’s ruling Marxist-
Leninist party mounted along with the financial bankruptcy of the state in 1989,
the authoritarian incumbents employed a tactic that had successfully been
used previously to quell social unrest and build consensus around the party’s
path forward: a National Conference. In February 1990, the ruling Parti de la
Révolution Populaire du Bénin (PRPB) called together an assembly of delegates
that was deliberately meant to be representative of all social, religious, profes-
sional, and political interest groups whose aim was ostensibly to chart the course
forward in support of the current regime (Nwajiaku 1994). Yet, on the second
day, this National Conference declared itself sovereign, proclaiming that all its
decisions would be legal and binding. This severed the control of the ruling party
from directly determining the institutional order of the new political system.
Surprisingly, the military and ruling party elites accepted the decision and con-
tinued to participate, largely because the bankruptcy of the state meant that the
status quo was not sustainable, and this process deliberately did not exclude
them entirely (Nzouankeu 1993).
The Conference put in place, through negotiations with all participants in-
cluding the military and representatives of the previous government, a constitu-
tional order and electoral institutions to usher in a competitive democracy. The
National Conference itself disbanded the sitting PRPB government but did not
banish its powerholders from the table in crafting the new order. The inclusivity
of the democratic bargain was premised on this foundation: “With the outcome
of any election uncertain at best, those who were likely to compete for power [and
have informal channels of access to it] found incentives to create a system that
would not guarantee absolute power to the elected but rather limit their rivals’
ability to monopolize the regime” (Magnusson and Clark 2005, 559). The goal of
the legislative branch with low barriers to entry was to avoid ethno-regional po-
litical dominance, by ensuring that parties and candidates representing different
regions of the country would all be included, no regionally based party would be
excluded, and access to resources and power would be distributed widely.
The trajectory of President Kérékou is also illustrative in this regard. As the
head of the ruling party and president, Kérékou oversaw the financial decline
and bankruptcy of the state. And the bankruptcy diminished the president’s
networks of his former patrimonial base, such that both the ruling party and the
military were simultaneously more autonomous from the president and weak-
ened in their overall capacity vis-à-vis the population and potential counter-
elites. The National Conference was conceived as way to build conciliatory
support to end the political and economic deadlock. When the body declared
itself sovereign, Kérékou lacked the support of the international community,
the military, and the loyalty of his core party elite as well as the civil administra-
tion of the state. Therefore, he had no choice but to accept the declaration. While
the ruling party was dismantled, the President was maintained as a figurehead
and allowed to stay on in his official capacity. This inclusivity was central to the
fragmented, discombobulated ruling elite’s acceptance of the new democratic
system. The president was assured a role in the proceedings, there was reconcili-
ation rather than retribution, and the ruling party was able to break up into new
constellations and be incorporated into new party configurations to continue to
vie for power.
Regarding this process of new rule formation in the National Conference, “it
is almost certainly a mistake to equate these with the greater democratic pro-
clivities of groups” represented (Nwajiaku 1994, 431). The crisis and resulting
political reformulation provided an opportunity for new associational groups to
form and a new arena for battles between the leaders of the fading ancien régime
and counter-elites who used the new political space to legitimize their claims to
power. Some actors were committed to democracy in their ideology, while others
were using the new structures to contest for power. Rising opposition elites were
instrumental in using the political reformulation to ensure their inclusion into
new power dispensations in the future and setting rules that would ensure their
future ability to compete and collude in that space (Banégas 2003). Low barriers
to entry for party formation and competition was one key element of the new
democratic order to facilitate inclusive bargaining of elite recirculation under
new party organizations (Riedl 2014).
Therefore, even at the point of the National Conference, democracy was not
the obvious outcome. President Kérékou’s maintenance as a figurehead and
ability to compete in the founding multiparty elections was critical to reassuring
the security forces and northern region that their interests would not be excluded
in the new regime. The military’s relative weakness due to the state’s bankruptcy
was also a key factor in allowing the National Conference to proceed to elections.
The military did not prop up the incumbent regime, nor were they sufficiently
coherent and organized to install a temporary military regime to guide the tran-
sition period and overly influence the outcome. Benin’s National Conference
was inclusive of the military, but neither they nor the PRPB elites were domi-
nant. They were sufficiently assuaged that their past transgressions would not
be prosecuted. And though the ruling party was conceding hegemonic control,
it was assured future access to power, and therefore, partial control. Democratic
elections and rules were the formal guarantees of this bargain, and they allowed
agreement on access and partial control.
Rising opposition elites included members of the legal establishment, busi-
ness representatives, and ex-patriots returning to the scene. In Benin, contrary
to South Africa, the opposition itself was not coherent or well formed from years
of liberation organization. The opposition elites in Benin were themselves fluid
and uncertain about their relationships to one another, and their individual
relationships to future structures of power (Diop 2006; Riedl 2014). This, too,
shaped the entire delegation’s preferences for inclusivity and participation of all
past and future political entities.
Their principal concern was to avoid the chaotic turmoil and crippling ethno-
regional divisions that had plagued the country following independence. To do
so, they sought to maintain President Kérékou as a figurehead, and they allowed
him and other members of the defunct PRPB to compete in the founding mul-
tiparty elections as independent candidates and under new party labels. In
shaping the new system, the emphasis was on inclusivity, consensus, represen-
tation of all forces of the nation, and achieving a stable regime system that could
address the deep crisis the country was facing (Nzouankeu 1993). The insta-
bility of the moment created incentives to ensure that all could participate in the
new political system, and to avoid shutting out a particular group. The National
Conference dialogues, the rising opposition elites, and the resulting institutions
were based on the premise that stability and recovery could best be achieved
through broad participation and inclusion. Opposition activists in exile were
given amnesty, and they returned to the country to “put the past behind us and
build the future. Everybody realized that we almost ended up in a civil war, and
that we had someone at the head of the country who realized that and adapted
his method in favor of peace, and that’s what counted” (author translation).11
A nonpartisan bishop who was respected for his neutrality and consensus-
seeking was selected by the Conference to preside over the agenda. This priority
shaped the institutional formation of political party registration, electoral code,
and the constitutional court that would help stabilize democracy over the next
three decades.
Benin’s democracy has endured, in large part, because the contestations over
elite access and ascension to power are waged as electoral, democratic, and con-
stitutional mobilization (Gisselquist 2008). The challenges have been many, with
the most extreme tests hinging on whether incumbent presidents would accept
defeat (after a re-election loss) or leave office after their second terms (in accord-
ance with term limits), and whether defeated challengers would accept electoral
results. The combination of domestic mobilization, elite power-sharing through
pre-and post-electoral bandwagon coalitions around the executive, and elec-
toral and judicial institutions maintained democracy through an explicit focus
on inclusion and consensus.
One of the informal mechanisms of successful inclusivity in Benin is the elec-
toral and political practice of grand coalition bandwagoning among almost all
candidates, parties, and representatives in the National Assembly, around the
president. The practice began with Nicéphore Soglo’s presidency (1991–96)
but was expanded greatly when he was defeated by the returning “chameleon”
Kérékou (nicknamed for his ability to reconfigure himself to the new demo-
cratic environment) in 1996 (Banégas 2003). Soglo built up a traditional political
party apparatus, the Renaissance Party of Benin (RB), and a coalition of six addi-
tional parties in the National Assembly organized around supporting his ruling
party supporters—but also an opposition set of parties. Kérékou ran as an inde-
pendent candidate, and his victory in 1996 was followed by a massive reorien-
tation of party alignments for the 1999 National Assembly elections. While the
RB maintained the plurality of seats (twenty-seven out of eighty-three), many of
the remaining fragmented party landscape joined in a presidential bandwagon
to support the new President Kérékou. This also provided autonomy for some
of the MPs and party leaders who sought to contest future presidential contests
as independents or on their own party labels. In the 2003 National Assembly
elections, the Presidential Movement coalition won a total of fifty-two of the
eighty-three seats, then composed of eight parties. This pattern continued with
successive presidents, giving maximum flexibility to the inclusive bargain of lim-
ited power-sharing that has served as a mechanism of democratic continuity
in Benin.
Low barriers to entry for parties and candidates to compete in elections have
also contributed to Benin’s democratic endurance through ensuring access
to political power to all regions and groups across the country. The prolifera-
tion of candidates and parties, and overall fluidity in the political party land-
scape, has encouraged grand coalitions by avoiding stark opposition camps,
allowing shifting allegiances, and ensuring that former authoritarian ruling
party members could be regenerated into new, viable elected representatives. No
matter what region or background, political elites could align with successive
presidents and have access to power.
In institutional terms, the National Conference prioritized the principle
of participation and wanted all groups to be easily represented—therefore,
both forming a party and running for election as an independent were facil-
itated through the electoral code and political party charter (Riedl 2014). The
resulting proliferation and fluidity of parties and independents further con-
tributed to the bandwagoning tendencies, as party labels and distinctions
had little meaning; party elites were free to ride the coattails of the presiden-
tial movement. In doing so, they maintained access to state resources and
supported the ongoing democratic project institutionally because they were
not cut out entirely.
The self- prescribed autonomy Benin’s Constitutional Court has from
the executive branch, combined with a democratic political culture among
the elite, which stems from the tradition established by the aforementioned
National Conference, has allowed for the Constitutional Court to greatly
contribute to democratic endurance (Bockelie 2013; Magnusson 2001). The
court has undergirded Benin’s democratic tradition since the transition
to democratic rule by giving citizens a channel through which they can file
complaints, increasing political transparency (Gisselquist 2008). The courts
greatly heightened their legitimacy during the critical 1994 budget crisis and
the 2006 elections. During the 1994 budget crisis, which emerged as a result
of an executive–legislative dispute, the Constitutional Court stepped in and
resolved the conflict, which marked the first time in Benin’s postcolonial his-
tory that a constitutional authority supplanted the military as the arbitrator
of institutional disputes.12 The courts also played a crucial role in the 2006
elections, which marked the first time that a sitting president had to abide by
a term limit. The courts, along with Electoral Commission, established key
procedures that would be followed for the elections and that effectively man-
aged the debates surrounding the constitutional age limit and term limits to
be upheld.
Critically, the Constitutional Court upheld the principle of consensus by
striking down a Constitutional reform that had passed the National Assembly,
declaring that though it had passed the necessary numerical threshold of votes,
because it did not have the support of all opposition deputies, it did not truly re-
spect social consensus. The Court’s action was perceived as a democratic check
on the legislative dominance of the presidential majority, upholding the princi-
ples of the National Conference of consensus and supporting the mass mobili-
zation of the public to safeguard the constitution from change (Ologou 2017).
This key moment for Benin’s democracy reinforced the values of inclusion and
the necessity of broad-based agreement to enact any institutional change. By
establishing their legitimacy, the courts were able to provide an alternative to
political strife. Citizens connected their protests and defense of the Constitution
with the Court’s decision. This affirmed public trust in the judicial branch, and
the sense of agency of the general population in protecting the nation’s demo-
cratic tradition.
The citizens’ mobilization and victory represent a form of continuity of the
principles of the National Conference itself. Participation and consensus were
embedded at the mass level, because the Conference itself included many social
representatives. Domestic mobilization has been important in times of dem-
ocratic challenge in Benin, particularly around questions of constitutional re-
form and presidential term limits. Benin’s constitution has been safeguarded
by domestic protests to defend the consensus principle and to maintain term
limits. After former authoritarian leader and reformed democrat President
Kérékou was elected and governed from 1996 to 2006, he considered whether
to try to amend the Constitution to allow a third term run. But a mass mobi-
lization campaign—supported financially and organizationally by opposition
elites—rallied around a “Don’t Touch My Constitution” campaign in the streets
and markets. Ultimately, Kérékou did not proceed with any formal attempt to
change the constitution or otherwise delay elections, and the 2006 elections
marked Benin’s third democratic alternation in power to outsider candidate
Boni Yayi. President Yayi again tested the question of flexible term limits in
2016, and the Constitutional Court maintained its vigilance. President Yayi also
built an inclusive coalition by integrating ministers and public servants from
across the regional divisions of the country into his administration. The elite
bargain of inclusivity, limiting hegemonic control, and allowing alternation
has been tested at the bounds but maintained by the relative dispersion of eco-
nomic and political power and the role of popular mobilization, supported by
counter-elites and channeled through institutions designed to limit and check
these tendencies.
Since President Talon was democratically elected in 2016, demonstrating yet
another successful executive turnover in Benin, the model of inclusive power-
sharing through grand coalitions has been attacked and undermined. Talon’s
control of power and resources in the economic sphere (the “King of Cotton,”
Talon controls the majority of Benin’s key export crop as well as management
over the port’s import-export revenue) facilitated his centralization of the polit-
ical sphere as well. He previously served as the financial backer of the plethora
of parties and candidates, and contributed to their fluidity. Once he decided to
enter the political sphere himself, there was no longer a need to support other
political candidates to assure his economic interests, and his goal became to
concentrate power and serve as a chief executive for the country. Talon has
instituted a series of political reforms that break down the previous foundations
of democratic endurance, including new appointees to the Constitutional
Court,13 a new political party registration code that has barred opposition
parties from participating in elections (notably beginning with the 2019 leg-
islative elections), and the creation of an anticorruption court, the Court of
Punishment of Economic Crimes and Terrorism (CRIET), which is widely ac-
cused of being politicized to attack the president’s political opponents (Freedom
House 2020a). These reforms raise red flags for democratic governance, human
rights, and effective accountability.
Talon’s more extreme exclusion of key political players in Benin has sev-
ered the compact of collusive power-sharing that undergirded Benin’s democratic
model since the transition to multiparty competition. The basis of the National
Conference provided a strong institutional foundation to maintain democratic
governance by prioritizing widespread access to state power, broad-based rep-
resentation, and political party fluidity to regenerate expired autocrats. But the
even more important legacy of the National Conference process was to create
an inclusive coalitional model of access to state resources and power among a
variety of regional and sectoral interest groups. Presidents Nicéphore Soglo
(1991–96) and Mathieu Kérékou (1996–2006) allowed broad bandwagon
coalitions to form around them in the National Assembly and in their admin-
istrative appointments. Boni Yayi (2006–16) maintained the practice with
some targeted closure, and Patrice Talon’s aggressive exclusion marks the end
of this compact, which previously facilitated political alternation and demo-
cratic endurance. The consequences of this closure include increased political
instability at the elite level, protests and citizen–state clashes, degradation of
democratic institutions to serve political centralization and autocratic ends,
and diminished mechanisms of accountability and inclusion for citizens across
the country.
International Powers and Normative Commitments

to Democracy: Alternative Arguments
Two common perceptions of democratization and endurance in sub-Saharan

Africa in general, and South Africa and Benin in particular, relate to the role
of international support for democracy and the role of committed democratic

ideologues.
As the analytic narratives of South Africa and Benin make clear, interna-
tional pressure intensified the existing domestic crisis—adding to the financial
pressures of the apartheid ruling class and the fiscal bankruptcy of the Beninese
state. This pressure made it likely that incumbents would stare down the prec-
ipice and realize the need for a change to the status quo. In South Africa, inter-
national factors were significant to democratization itself in two distinct ways.
First, the dissolution of the Soviet Union and the end of the Cold War meant
the ANC was no longer a communist threat, and therefore was a viable nego-
tiating partner in Western eyes, as well as to the domestic ruling party. The
ANC had to shift their own strategy to be more procapitalist and promarket;
they tempered their means (violence) and ends (reduced their focus on re-
distribution and adapted to a more neoliberal orientation). The transition
mentality coming out of Eastern Europe with the end of the Cold War also
reshaped political leadership within the National Party and the type of negoti-
ated settlements the party might be willing to consider in response to the rising
threat of democratic majoritarianism (Saunders 2009). Second, foreign aid fol-
lowing the transition helped buttress the new inclusive power-sharing coali-
tion, with a massive inflow to expand public services to the general population.
These factors influenced the structural constraints and opportunities that both
the authoritarian incumbents and rising opposition faced before and after the
transition, but they did not predetermine the content of the negotiations or ne-
cessitate their survival.
Similarly in Benin, the international context created external and in-
ternal pressures for economic and political reform that pushed the author-
itarian incumbent to face the need for some kind of transition (with the
dissolution of Soviet support), but it was not determinative of the outcome.
Neighboring examples in Togo, Burkina Faso, and Côte d’Ivoire make clear
that a transition to multiparty elections without ceding power and some
neoliberal economic reform was enough to satisfy Western donors (Gazibo
2013). French involvement in the transition in its former colony created
conflicts of interest with the Beninese population, as French pressures for
the transition process to unfold in a controlled manner seemed designed
to guarantee the authoritarian incumbent’s maintained position (Nwajiaku
1994). Yet France did push for economic and political reform in Benin,
because President Kérékou had been more isolated from Western de-
velopment assistance over the past decade, and France withheld funds
earmarked for relief of the economic crisis until the National Conference
was held (Seely 2009). Following the transition, USAID and other Western
donors have largely focused aid on the health and education sector, because
they could take for granted the self-sustaining democratic foundations

Benin developed (McMahon 2002). As in South Africa, Benin’s public ser-
vice sector has benefited from these donor relations, which indirectly has
helped to maintain the inclusive coalition at the political elite level. But the
rapid autocratic decline that President Talon invoked demonstrates how
little influence donor calls for maintaining democracy have in the face of
domestic will to centralize power. The bargain that sustained Benin’s de-
mocracy for decades was one of domestic inclusion.
The role of democratic ideologues plays a similar role to international aid
in South Africa and Benin: facilitating the conditions that make a democratic
bargain possible but not guaranteeing its success or endurance. In South Africa,
Nelson Mandela’s iconic status is legendary. His role as a broker in the negoti-
ating process was crucial to forging consensus. But Mandela was released from
prison and allowed to participate in negotiations because of the changing struc-
tural conditions that forced the incumbent party to reckon with their uncertain
futures. Mandela came to the helm of the ANC with a centrist position because
of the changing structural conditions that brought together the labor union, the
Communist Party, and the militant “Spear of the Nation” wing to cohere behind
a negotiated transition process. Mandela was part of a movement that shaped
his own ideology and strategy over time, as well as those of his oppressors. And
the movement, and the institutions forged through compromise, continue to
sustain South Africa’s democracy through party leadership changes and gener-
ational changes.
In Benin, there was no parallel movement, as the fluidity of the transition
came out of diverse groups representing different sectors of society. And there
was no parallel symbolic figurehead; neutral interventions were sought by non-
partisan lawyers and religious leaders to steer the dialogue and help craft the
new institutional foundations. In Benin, many of the key players were not com-
mitted democratic ideologues, given the representation allotted to the former
ruling party and military elites. They were often self-interested and embattled
elites staring down the precipice and opting for institutional compromise given
the infeasibility of maintaining the status quo. The fact that democratically ori-
ented meditators could play a key role was a consequence of once hegemonic
elites, and their opponents for power, attempting an institutional reform after
prior decades of recalcitrance and rising conflict. Successive presidents flirted
with democratic abrogation, particularly Kérékou and Yayi both considering
constitutional changes and third terms, but the fluid alliance of economic and
political supporters with their ability to mobilize civil society protests, made
these men reconsider their strategic course of action and cede power to the next
elected executive.
Conclusion
The cases of Benin and South Africa defy scientific predictions regarding the
emergence and durability of democracy because of the barriers to achieving si-
multaneous stability, accountable competition, and inclusion in states that were
never crafted on that premise. The bandwagon model in Benin and the domi-
nance of the ANC coalition in coordination with remnants of the ancient apart-
heid regime in South Africa provide guarantees for representation, and mitigate
against hegemonic control. This represents the coordination among different
factions within the broader society, and the shared interests provide a foundation
for continuing the democratic bargain.
The surprising relative success of these two countries lies in their shared am-
bition to avoid economic redistribution and retributive justice but rather to seek
reconciliation and an integrated political future. Ruling elites were forced to
reckon with their declining ability to maintain hegemonic control. In the face
of potential civil war (revolutionary, ethnic) or a military insurrection and re-
placement with the next authoritarian regime, the ruling political class sought to
ensure protections, partial political influence, and their ability to continue accu-
mulation. The strength of the South African institutional environment provided
a foundation for that negotiation, and the National Conference in Benin swept
the old ruling party out but allowed them to reenter in new forms.
The critical factors thus include institutions that facilitated moderation and
provided the opportunity for negotiation and inclusion, such as the party organi-
zation of the ANC, the selection of proportional representation to allow multiple
representatives in South Africa, and the National Conference model and political
party charter in Benin. The successful outcome meant installing a viable multi-
party competition by virtue of the “institutionally oriented” strategies that political
elites deployed across the transition to share power. The return of certain former
politicians, and political parties, shaped the politics reducing the uncertainty of
competition, a factor that has reinforced the institutionalization of power.
Moreover, these transitions and their enduring institutions interact through
a historical process at a particular moment of world-historical time (of ideolog-
ical and material pressures for multiparty liberal democracy) and in relation to a
set of structural factors that determines the distribution of social and economic
power within highly cleaved societies. In Benin and South Africa, uniquely, these
allowed a preference for partial control rather than hegemonic control. The ex-
panded set of political elites facilitated economic accumulation among broader
constituencies, political stability, and increased physical security. In large part, it
was the depth of the past cleavages and fear of how they could otherwise unfold
that made this bargain more valuable and sustainable.
Notes
1. The legacy of anti-colonial liberation movements and post-independence nation-

alist models created deep ideological, organizational, and strategic ties to the Soviet
Union across parts of the continent.
2. Or, more generally, as the ruling class becomes increasingly incapable of repressing
the discontent and demands for inclusion.
3. As the economic situation in South Africa grew dire in the 1980s, the predominantly
white and wealthy business owners realized that some form of negotiated settlement
with the opposition would be necessary in order to create an environment that was
favorable once again for businesses (Marais and Davies 2015, 4–6).
4. Apartheid was officially instated in South Africa in 1948, when the Afrikaner-based
National Party (NP) won political power.
5. Bureau of Intelligence Research, South African Labor and the Wiehahn Reforms, re-
port no. 1290, December 12, 1979, NSA Archives, p. 12.
6. Director of Central Intelligence, “South Africa: The Politics of Racial Reform.”
Interagency Intelligence Memorandum, January 1981, p. 2.
7. Crucially, the fact that the ANC push for majoritarian rule was not focused on ex-
propriation or complete exclusion of the NP and white minority more generally, and
their increasingly conciliatory approach, fostered the deep splits within the white
ruling coalition, between the British and Afrikaners, and even within the Afrikaner
support for apartheid. The pressures created by sanctions and the rising mobiliza-
tion of the liberation movement meant that it was becoming unsustainable for the
ruling party to maintain its hegemonic political control, and this forced discussion
of reforms. Once reforms were on the table, increasing splits within the ruling party
and popular protest expressing discontent pushed an increasing number of citizens
to mobilize against the regime.
8. Director of Central Intelligence, “Prospects for South Africa: Stability, Reform and
Violence,” Special National Intelligence Estimate, SNIE 73.2–85, August 1985, p. 3.
9. The removal of sanctions and accompanying increased flexibility in production,
increased human resources development, and increased involvement of employees
in managerial decision-making stand out as drivers of economic growth and stability
that accompanied the post-apartheid years.
10. For example, the para- statals, contracts, licensing decisions, and commercial
networks that controlled profitable informal trade were main drivers of economic op-
portunity (Adjaho 1992; Chabi 1993).
11. Judicaël Zohoun, “Me. Robert Dossou apprécie les acquis et les avancées.” [Mr.
Robert Dossou appreciates the gains and advances.] 24 Heures au Bénin, February
24, 2020, https://www.24haubenin.info/?Me-Robert-Dossou-apprecie-les-acquis-et-
les-avancees.
12. Throughout the first term of the democratically elected president, Nicéphore Soglo,
multiple institutional crises plagued the regime. An environment of uncertainty
was created by arms theft and low-level military discontents as well as political
confrontations between the president and the National Assembly and an assertive
new Constitutional Court (Magnusson and Clark 2005). The military, crucially, did
not intervene in these disputes. The Constitutional Court emerged as a legitimate and
respected public institution, one that helped minimize the centralizing tendencies of
executive power and the potential overreach of the National Assembly, and ultimately
forced mediation.
13. In 2018, Talon made his personal lawyer, Joseph Djogbénou, President of the
Constitutional Court and has in general greatly imposed his influence on the Court,
thus reducing the Court’s objectivity (Kohnert and Preuss 2019; Stroh 2018).
5
Georgia, Moldova, and Ukraine
Democratic Moments in the Former Soviet Union
Lucan Ahmad Way
After the collapse of the Soviet Union in 1991, post-Soviet countries1 confronted
extraordinarily inhospitable conditions for democratic development.
Totalitarian rule had decimated all independent civil society that might chal-
lenge state power and left post-Soviet countries without powerful or autonomous
private sectors. Furthermore, in contrast to post-communist states in Central
Europe and the Baltics, post-Soviet states were never given an opportunity to
join the European Union—a fact that deprived elites in those countries of critical
incentives to democratize. Partly as a result, the region failed to produce a single
stable democracy. Regime types ranged from quasi-totalitarianism (such as in
Turkmenistan), to classic authoritarianism (such as in Russia and Azerbaijan), to
the competitive authoritarianism we observe in such countries as Armenia and
Kyrgyzstan.
Nevertheless, an exclusive focus on the region’s record of democratic failure
obscures important cases of democracy in hard places. Even in such inhospi-
table terrain, certain countries have managed to attain democracy—albeit for a
relatively short period of time. This chapter focuses on “democratic moments,”
a term I use to describe cases in which democracy (or something very close to
it) emerges but remains unconsolidated and short-lived.2 During these demo-
cratic periods, countries host free and fair elections, allow the existence of free
media, and (very often) witness peaceful turnovers of power via the ballot box.
Examples include Georgia in the 2010s; Moldova in the 1990s and 2010s; and
Ukraine in the early 1990s, late 2000s, and late 2010s.
A comparison of Georgia, Moldova, and Ukraine suggests that “democratic
moments” in hard places are a product of three mechanisms: First, authoritarian
weakness results in the establishment of “pluralism by default.” In such cases, plu-
ralism emerges not from a particularly strong civil society, powerful institutions,
or democratically minded leaders but instead from weak control over the co-
ercive apparatus and underdeveloped ruling parties. Would-be autocrats have
been unable to rely on armed forces to repress opposition, or to use ruling parties
to keep allies in line. A second factor is the persistence of media and political
Georgia, Moldova, and Ukraine 129
parties from the prior authoritarian period. Although these institutions once
provided key support for authoritarian rule in the previous regime, they may
help uphold fledgling democratic moments by creating checks on the new re-
gime. Finally, a third factor—Russian influence—has had both negative and
positive effects on democracy. On one hand, President Vladimir Putin has ac-
tively promoted autocratic behavior of pro-Russian leaders in these countries,
and threats from Russia have sometimes encouraged media restrictions. At the
same time, Russian presence in the region has made anti-Russian leaders in these
countries highly susceptible to even mild international democratizing pressure.
While international pressure was much less powerful and consistent in the post-
Soviet region than in Central Europe, it nonetheless encouraged the emergence
of democratic moments by preventing certain power-holders from indulging in
the most serious abuses of democratic norms. Overall, these three factors have
promoted often very real but unstable democratic development.
In the following section, I discuss the obstacles facing democracy in the former
Soviet Union and explicate the three factors described above. Then, I explore
how these factors worked to produce democratic moments in Georgia, Moldova,
and Ukraine—the three most democratic countries in the former Soviet Union.
Both Moldova and Ukraine experienced periods of genuinely democratic rule in
the post–Cold War era. While Georgia failed to fully democratize, the govern-
ment maintained democratic contestation.
The Former Soviet Union as a Hard Place
In the 1990s, many observers assumed that the twelve newly independent states
were well on their way to democracy.3 However, seventy years of communist
rule had generated highly unfavorable conditions for democracy. The Soviet
totalitarian system undermined democratic development by politicizing the
state bureaucracy and preventing the emergence of any economic or political
forces outside of state control. In contrast to authoritarian regimes elsewhere, in
which an authoritarian executive coexisted with a relatively autonomous private
sector, Soviet totalitarian rule had maintained direct control over the economy
and almost all aspect of people’s lives. Until the late 1980s, economic activity—
from large industrial plants to street cafes—was directed by a vast party-state
bureaucracy.
The Party prohibited the emergence of any political organization or activity
outside regime control. Independent parties, trade unions, or even nonpolit-
ical associations such as sports leagues or chess clubs were strictly forbidden.
Independent political initiative was strongly discouraged, even if it was in osten-
sible support of the regime. Thus, when a thirteen-year-old Marina Morozova,
a friend of the author, gathered four of her friends to protest American nuclear
weapons in front of the United States embassy in Moscow in the summer of 1978,
this endeavor was met with tremendous suspicion by Soviet authorities. A KGB
officer who questioned the girls explained, “Today, you are organizing a protest
against the United States. But tomorrow you could protest against the Party!”
These characteristics of the Soviet system meant that the twelve independent
states emerged in 1992 without any established political or economic organiza-
tions or networks that could challenge the government.
Simultaneously, the pervasive partisan penetration of state institutions such
as the courts and police meant that such institutions were highly vulnerable to
manipulation in the post-Soviet era. As a result, new governments faced rela-
tively few institutional obstacles in using tax authorities and coercive and regula-
tory agencies as weapons to put down opposition. In the decades after 1991, tax
officials, for example, targeted opposition politicians and businesses with audits
and fines not applied to friendly forces.
Finally, post-Soviet countries confronted relatively weak external democra-
tizing pressures. In contrast to their Central European counterparts, autocrats
in the former Soviet Union met with softer and more sporadic external demands
for democratic change (Kopstein and Reilly 2000; Levitsky and Way 2010; Levitz
and Pop-Eleches 2010). Most importantly, not a single post-Soviet country was
given the opportunity to join the European Union, which emerged in the 1990s
as one of history’s most effective democracy promoters (Vachudova 2005). Aid
was only loosely tied to democracy and, where pressure was applied, it was often
half-hearted.4 As a result, autocrats did not face any kind of clear incentive to re-
frain from abuse.
Such conditions contributed to the overwhelming failure of democracy
in the region. In 2019, seven of twelve post-Soviet countries were ranked by
Freedom House as “not free,” while the rest were labeled “partly free.” The only
post-Soviet country ever to have been ranked as “Free” by Freedom House was
Ukraine between 2005 and 2009. Of the seventy-six executive elections held
in the former Soviet Union between 1991 and 2019, only eight have resulted
in the peaceful democratic transfer of power to opposition forces (Freedom
House 2019a).5
Democratic Moments in the Former Soviet Union
This litany of failure nonetheless conceals important democratic successes in the

region. In addition to Ukraine, two other countries—Georgia and Moldova—
experienced democracy or something very close to it after the Cold War. During
democratic moments, elections were generally free, fair, and competitive; all
adults had suffrage; there were few if any violations of freedom of speech, press,
and association; and power was held by elected officials rather than by non-
elected “tutelary” authorities (e.g., militaries, monarchies, or religious bodies)
(Dahl 1971). At the same time, democracy was not consolidated: leaders regu-
larly tried to abuse their power and change the fundamental rules of the game.
Democracy was not regularized or taken for granted and was often short-lived.
Nonetheless, the emergence of genuine democracy in such inhospitable terrain
cries out for explanation.
Georgia, Moldova, and Ukraine represent the three most democratic coun-
tries in the former Soviet Union according to V-Dem (Coppedge et al. 2021).
Detailed case study analysis (see Way 2015) suggests that significant democratic
moments emerged in Georgia in the mid-and late 2010s; in Moldova in the 1990s
and again in the 2010s; and in Ukraine in the early 1990s, late 2000s, and finally
after 2014. While this description of regime evolution does not precisely match
V-Dem’s scorings over time, it represents the author’s best assessment based on a
detailed analysis of the cases.6
First, Moldova quickly transitioned to democracy after the Soviet collapse and
witnessed peaceful democratic turnovers in 1996 and 2001. However, the elec-
tion of the Communist Party of Moldova in 2001 led to a clear slide into com-
petitive authoritarianism after the government began a generalized assault on
media freedom. Following the ouster of the Communists in 2009, the country
witnessed a resurgence of democracy and peaceful democratic turnovers in 2016
and 2020.
Next, Ukraine came very close to democracy in the early 1990s but witnessed
backsliding in the late 1990s. It then experienced a clear democratic moment
following an electoral revolution in 2004 (the Orange Revolution) that resulted
in the election of Viktor Yushchenko. Under Yushchenko (2005–2010), Ukraine
was a full democracy: there were no serious assaults on media or other violations
of civil liberties, the opposition was given the fullest freedom to organize, and in
2010 Yushchenko left office after a completely free and fair election that saw the
democratic election of Viktor Yanukovych. The democratic moment ended after
Yanukovych jailed opposition leaders and undermined media freedom. But after
Yanukovych’s ouster in 2014, the country again came close to democracy and
witnessed a peaceful democratic turnover in 2019.
Finally, Georgia, which experienced total state breakdown in the early 1990s
and had weaker conditions for democracy than either Moldova or Ukraine,
experienced a particularly surprising democratic moment in the early 2010s.
In 2013, the country experienced its first peaceful and democratic turn-
over when President Mikheil Saakashvili left power in favor of the Georgian
Dream coalition. From 2013 until 2020, Georgian Dream governed in a rel-
atively democratic manner. The opposition was allowed to mobilize and the
quality of elections improved over time. However, in early 2021, the govern-
ment arrested the main opposition leader, marking a sharp turn toward greater
authoritarianism.
Where Do Democratic Moments Come From?
What explains the emergence of democratic moments? How are countries able to
overcome inhospitable conditions to become democratic—even for a relatively
short period of time? One possibility is that these cases emerge out of completely
stochastic or random variation. For example, such periods might result from the
unpredictable and idiosyncratic appearance of democratically minded leaders
who act in democracy-sustaining ways despite incentives to the contrary.7 Yet,
there are two reasons to doubt this explanation in the post-Soviet context. First,
it is hard to argue that democratic moments are randomly distributed across the
former Soviet Union—which is what we would expect if a stochastic factor such
as leadership were involved. Most obviously, no post-Soviet country with ac-
cess to significant natural resources experienced a democratic moment (Bellin
2004, 2012).8 All five post-Soviet cases that rely heavily on natural resources—
Azerbaijan, Kazakhstan, Russia, Turkmenistan, and Uzbekistan—ultimately cre-
ated quite stable and closed authoritarian regimes.9 (At the same time, as we will
see below, the transition between democracy and competitive authoritarianism
within Georgia, Moldova, and Ukraine was arguably more contingent.) Second,
leaders during democratic moments were not especially democratic, and they
engaged in repeated efforts to abuse democratic norms. This chapter contends
that democratic moments emerge less from these random sources of variation
and more from the systematic contextual factors of authoritarian weakness, the
persistence of old regime political forces, and geopolitical vulnerability. I discuss
each of these in turn.
Authoritarian Weakness
As I have argued elsewhere (Way 2015), democratic moments often result not
from the strength of democrats but from the weakness of would-be autocrats.
In such cases, autocrats have been too enfeebled to monopolize political con-
trol, steal elections, repress opposition, or keep allies in line—resulting in a
dynamic and genuinely competitive political environment. In particular, I high-
light two forms of authoritarian weakness that lead to “pluralism by default” that
undergirds democratic moments in the former Soviet Union: weak ruling parties
and authoritarian state apparatuses.
Authoritarian regimes with well-established ruling parties have historically

been seen as more durable than military or other types of nondemocratic regimes
(Brownlee 2007; Geddes 1999; Huntington 1968; Magaloni 2006, 2008; Svolik
2012). In turn, weak and poorly institutionalized ruling parties often promote
more dynamic political competition. Party weakness—defined by the absence
of a single, well-institutionalized party structure10—facilitates pluralism by en-
couraging elite defection to the opposition. Indeed, many opposition leaders in
the former Soviet Union—Leonid Kuchma and Viktor Yushchenko in Ukraine,
Mikheil Saakashvili in Georgia, Petru Lucinschi in Moldova—emerged out of
previous governments that were not governed by a single party.
Party weakness was particularly endemic in the post-Soviet context because
the Soviet collapse had resulted in the widespread destruction of older party
structures. Fearful that the Communist Party would undermine Perestroika
reforms, Gorbachev, as the Party’s leader, dismantled key party institutions and
hampered efforts by lower-level officials to defend party power in the face of the
first competitive elections in 1989 and 1990 (Way 2015, chap. 2). Furthermore,
the Party was widely blamed for the Soviet Union’s disastrous economic condi-
tion in the late 1980s and early 1990s. Thus, after the failed coup of August 1991,
the Party was banned in a number of new states, including Georgia, Moldova,
and Ukraine. In Ukraine, for example, party cells were dissolved and officials sud-
denly looked on party membership as a “contagious disease.”11 In some parts of
the former Soviet Union where party-state institutions remained relatively intact
or where leaders benefited from easy natural resource rents (e.g., Turkmenistan,
Kazakhstan, Russia, Azerbaijan), officials were quickly able to adapt. But in other
cases, such as Georgia, Kyrgyzstan, Moldova, and Ukraine, which lacked such
resources and witnessed significant breakdown of older institutions, leaders had
a much harder time reorienting and consolidating control.
Similarly, democratic moments may occur when authoritarian state power—
defined as the willingness and capacity of repressive and other state agencies to
support incumbent power—is limited.12 An extensive and cohesive coercive ap-
paratus is necessary to enable rulers to monitor, intimidate, and suppress po-
tential sources of opposition, and to manipulate elections. In another era, an
absence of such instruments might well have led to a military or civilian coup.
But in the post–Cold War era, coups are much rarer, and when they do occur,
they have most often been followed by democratic elections (Marinov and
Goemans 2013). In cases of weak repressive capacity, leaders have faced diffi-
culties carrying out any large-scale coordinated authoritarian action and have
therefore been vulnerable to even weak opposition challenges. Excepting those
cases in which weakness has led to a breakdown in social order, weak coercive
capacity has often provided key political opportunities for opposition mobiliza-
tion and often underlies “people power.”13 This was the case in both Ukraine and
Moldova in the 1990s, when chronically underfunded coercive apparatuses with

large wage arrears proved unable and unwilling to control challengers to incum-
bent autocrats, resulting in political openings that lasted for several years.
To be clear, even weak autocrats can engage in sporadic harassment of the
opposition. Pluralism by default certainly does not guarantee democracy. The
transition between democracy and competitive authoritarianism was sometimes
quite contingent. However, by hampering efforts to consistently suppress oppo-
sition and prevent elite defection, party and state weakness makes it much harder
for leaders to eliminate political competition—thereby generating frequent po-
litical openings and propitious conditions for democratic moments.
Persistence of Old Regime Forces
Next, I argue that the persistence of old regime political forces—in particular,
political parties and media—may facilitate democratization. The survival of old
regime parties and press can create checks on the power of new governments.
James Loxton (2018, 28–29) has argued that former authoritarian ruling parties
can contribute to party system institutionalization by anchoring one side of
the regime cleavage. The checks imposed by autocratic successor parties and
institutions are particularly important in the post-Soviet context, where leaders
have faced few organized challenges to their rule. The total disintegration of
old regime organizations undermines democracy by depriving the opposition
of potentially powerful tools to challenge the new government. For example, in
Georgia, the successive and complete destruction of old regime forces in 1991,
1992, and 2003 left new governments with a freer hand to abuse power. Political
forces that might have provided democratic challenges to new governments in
elections or via opposition media were absent. By contrast, in Ukraine in the
late 2000s and Georgia in the 2010s, the persistence of media and parties from
the previous government created checks on government power—thereby bol-
stering pluralism.
Vulnerability to External Pressure
Finally, the post-Soviet space is uniquely shaped by the regional dominance of

Russian autocracy. In fact, all three countries covered here suffered Russian mil-
itary intervention. Many existing accounts emphasize the ways in which Russian
presence promotes autocracy in the region by strengthening pro-Russian author-
itarian forces such as Leonid Kuchma and Viktor Yanukovych in Ukraine and the
Communist Party in Moldova. At the same time, when explicitly anti-Russian
forces have come to power, Russia has undermined efforts to consolidate authori-
tarian control by financing opposition forces and making incumbents the targets
of Russian-generated opposition propaganda (Way 2016). Hostility from Russia
has also enhanced dependence on Western support—thereby strengthening
the impact of Western democracy promotion. Georgia, Moldova, and Ukraine
have all been highly dependent on Western aid in the face of Russian military
aggression. External vulnerability made these countries more susceptible to
Western democratizing pressure. Such pressure was much weaker and more spo-
radic than that created by the European Union in Central Europe (Levitsky and
Way 2010), but could still generate democratic moments. Most significantly, in
Georgia in 2013, such pressure created conditions for the country’s first peaceful
democratic turnover in power. Overall, the contradictory impact of Russian hos-
tility has likely enhanced regime instability in these countries.
In the next section, I describe how the mechanisms described above produced
democratic moments in Moldova, Ukraine, and Georgia. All three countries
emerged out of at least a half-century of totalitarian rule, with relatively weak
civil societies and state apparatuses vulnerable to partisan capture and manipu-
lation. Yet each enjoyed extended periods of political competition that, while not
consolidated, was real and often genuinely democratic. I find that these demo-
cratic moments emerged from different configurations of the causes described
above. In Georgia, Moldova, and Ukraine, weak ruling parties were unable to
erect new authoritarian regimes. In Ukraine and Georgia, a major obstacle to the
incumbent’s potentially authoritarian designs was posed by the persistence of
parties and media associated with the prior regime. In Georgia, democratization
was further aided by its vulnerability to international democratizing pressure,
which imposed limits on how far incumbents could go to cement power.
Moldova
While Moldova had weak democratic prerequisites, it experienced two demo-

cratic moments—in the 1990s and again in the 2010s—that were driven to an
important extent by authoritarian state and ruling party weakness. A poverty-
stricken, rural country with no serious Soviet-era dissident movement or pre-
Soviet democratic history, Moldova lacked any qualities scholars would associate
with democratic success when it became independent in 1991 (Crowther 1991,
184; Way 2015). Moldova was beset by ethno-linguistic tensions between a
Moldovan-speaking west and Russophone east. In 1989, Moldovans took to the
streets to protest Soviet-era Russification policies, with the result that Moldovan
was made the official state language. This, in turn, sparked opposition among
Russian speakers in the east, and a small-scale civil war ensued for several
months in 1992. In short, at independence, Moldova was not a likely candidate

for democracy.
Yet, Moldova in the 1990s experienced full democracy. Civil liberties were
respected. Media were free and the opposition was allowed to mobilize. While
President Mircea Snegur passed a draconian law in 1991 that prohibited criticism
of the president, the law was never enforced (FBIS 1992). In fact, in presidential
elections five years later, government media tended to be biased against the in-
cumbent Snegur (ODIHR 1996, 5). When Snegur lost that election to Speaker
of the Parliament Petru Lucinschi, there were no allegations of ballot stuffing or
other serious violations, and a peaceful transfer of power ensued. Lucinschi in
turn lost the presidency in 2001 in free and fair elections that “met international
standards for democratic elections” (ODIHR 2001, 1).
The rapid emergence of democracy in such an inhospitable context can, to
an important extent, be understood as the product of state and ruling party
weakness. First, the Moldovan state lacked serious coercive capacity. In the
early 1990s, the country’s armed forces consisted of “lightly armed” and un-
derpaid policemen, a “hastily assembled” army, and nationalist volunteers with
farm implements who were mobilized to fight separatists in the east during the
brief civil war of 1992.14 Discipline was also compromised by large-scale wage
arrears and pervasive low morale (King 2000, 192–93; March and Herd 2006,
365). The ruling party was similarly weak. As in other post-Soviet countries, the
Communist Party completely disintegrated after the failed Soviet coup of August
1991, which in turn led to the collapse of the Soviet Union. In the early 1990s,
old Communist forces attempted to regroup into the Agrarian Democratic
Party (ADP), a loose, nonideological alliance of state farm directors that initially
supported President Snegur (Socor 1992).15 The ADP, described as a “collection
of regional fiefdoms and personal cliques,” lacked a developed organization or
political identity and failed to survive beyond a single election in 1994.16 In 1994,
the party captured 56 of 104 seats in parliament but began to fragment within just
over a year. In the 1998 parliamentary elections, the ADP won no seats in par-
liament and promptly disappeared. A strong ruling party also failed to emerge
during the administration of President Lucinschi (1997–2001). Fearing the rise
of a competitor, Lucinschi sought to prevent the institutionalization of any single
political faction and ruled without the support of any party.17
The 1996 presidential election pitted three former supporters of the ADP
against one another: Snegur, Prime Minister Andre Sangheli, and parliamen-
tary speaker Petru Lucinschi. (Moldova was a semi-presidential system between
1991 and 2001 in which the president was considered most powerful.) By 1995,
the state was torn asunder by a tug-of-war between these “three whales,”18 who
each sought to use “their” part of the state against the others. This resulted, first,
in highly contested control over the already-weak coercive apparatus, which
made it hard for any single side to use force to obtain control. Simultaneously,
the head of the state media, Adrian Usatii, whom Snegur had appointed in 1989,
had a falling out with Snegur in 1995 and openly backed Lucinschi in the 1996
election.19 As a result, even as an incumbent, Snegur received generally nega-
tive coverage on state television—the only television to reach many rural areas
(ODIHR 1996, 7). Snegur also faced significant difficulties controlling local
governments—a fact that largely precluded serious vote fraud. In the end, Snegur
lost to Lucinschi 46 percent to 54 percent.
A former Soviet Politburo member adept at backroom politics, Lucinschi
(1997–2001) was seemingly well placed to monopolize political control. His allies
gained increasing control of parliament. Just after his election, many ADP dep-
uties went over to the president,20 and a pro-presidential coalition, the Bloc for
a Democratic and Prosperous Moldova (BDPM), gained a substantial number
of seats in elections in 1998.21 Allies of the president constituted 61 of 101 seats
in the legislature. On April 21, Lucinschi’s close ally, Dumitru Diacov, who had
worked with him since 1995, was elected speaker of parliament.22 Observers
widely expected Lucinschi to dominate the legislature (Roper 2001, 6).
Yet coalition weakness fundamentally undermined Lucinschi’s efforts to con-
centrate political power. The problem was not that Lucinschi lacked sufficient
allies in the legislature but that, in the absence of a single well-institutionalized
party, the president quickly lost control over the supporters he had. Fearing the
rise of a competitor to his power, Lucinschi engaged in divide-and-rule tactics,
seeking to disperse his support across a wide range of political organizations and
prevent the concentration of power in any single group.23 Thus, while the BDPM
had been created as a pro-presidential coalition, Lucinschi openly promoted a
number of competing groupings and independent deputies in the 1998 parlia-
mentary elections—a fact that alienated many of his allies and undermined the
electoral chances of the coalition.24 Lucinschi’s behavior may have been moti-
vated by a rational desire not to tie his political fortunes to a new group that was
uncertain to succeed; but his actions contributed to the perception that he was a
“bad-payer of political bills.”25
As a result, despite his strong starting point, Lucinschi was unable to con-
centrate political power. To begin with, the legislature deprived the president of
control over the media. When state television began to attack deputies deemed
critical of the president, parliament fired the head of state media and reduced the
president’s formal role in choosing his successor.26 Subsequently, control over
different parts of the government media complex was distributed among dif-
ferent parliamentary factions—which promoted pluralism in the media.27
Most critically, Lucinschi’s efforts to strengthen presidential rule were
overturned by those closest to him. Beginning in February 1999, Lucinschi
proposed to dramatically increase the power of the president.28 In response to
severe opposition from the legislature, including Diacov and the BDPM (which,
ironically, had run on a platform of strengthening presidential power), Lucinschi
organized a national referendum, which garnered majority support but failed
to attract sufficient turnout to make it binding (Crowther 2007, 276). While
Lucinschi actively sought Western support, the Venice Commission, a European
advisory body on constitutional law, concluded that the proposed changes were
“contrary to European democratic principles.”29 The president’s former allies, in
fact, began a campaign to eliminate the directly elected presidency altogether.
Diacov was not especially supportive of the parliamentary system, but he consid-
ered Lucinschi to be “totally unreliable” and concluded that strengthening par-
liamentary power was the only way to defeat him.30 In 2001, Moldova became a
parliamentary republic in which a president was elected by the legislature.
What explains Lucinschi’s self- destructive failure to build a party? His
mistakes cannot be attributed to political inexperience. A former member of
the Soviet Politburo and former first secretary of the Soviet Communist Party
of Moldova, Lucinschi was among the most seasoned politicians in Moldova.
When the Soviet Union collapsed, he adroitly pivoted away from the Communist
Party, which was banned in 1991, and quickly reemerged as the speaker of the
Moldovan Parliament. However, his abandonment of the Party was a double-
edged sword. In the immediate aftermath of the collapse, it allowed him to
quickly regain authority amid widespread anti-communism. However, in the
medium term, it left him without an organized base of support and with no ide-
ology and few organizational networks to build a new one. While not impos-
sible, building a new party in such a context was highly challenging. Like Boris
Yeltsin in Russia and Ukraine’s Kravchuk, who also retained power by rejecting
the Soviet Communist Party, Lucinschi confronted a bit of a catch-22 in the
1990s. On one hand, new political forces were too weak to rely on. On the other
hand, efforts to diversify allies away from these forces often alienated existing
supporters. As a result, Lucinschi, like other leaders during that time, chose ini-
tially to rely on his own personal power rather than invest in party building. The
result was pluralism by default.
The importance of ruling party weakness in facilitating pluralism is further
evidenced by Moldova’s regime trajectory after the 2001 elections that resulted in
a sizable victory (71 of 101 seats) by a much better organized Communist Party
of Moldova. Banned in 1991, the Communist Party had been revived in 1993
from the remnants of the old Soviet party by second-tier Soviet-era leaders who
embraced party symbols and ideology. As a result, they were able to quickly re-
build a relatively robust party organization and maintained strong discipline
under the leadership of Vladimir Voronin. While parliamentary rule is typically
associated with greater democracy, it allowed Voronin and the party to rule with
virtually no checks on their power. During the party’s tenure in power, most
policy decisions were made outside the legislature by the party hierarchy. In turn,
the rise of the Communists led to a notable decline in civil liberties and media
freedom. State-run television was censored and grew increasingly biased, and
independent talk shows were taken off the air (ODIHR 2005, 1; Way 2002, 131).
The Communist Party also was also much more effective than Snegur had been
at monopolizing control over the state in the 2005 and 2009 elections. As a result,
Moldova’s democratic moment was decisively brought to an end.
However, democracy reemerged in 2009 when the Communists were over-
thrown in the wake of post-election protests in April. The election left the Party
just one vote shy of the sixty-one votes needed to elect a president. Due in part
to a highly polarized atmosphere created by protest violence, the loose coalition
of opposition parties was remarkably able to deprive the Communists of that
single vote, and new elections were held.31 After repeat elections in July, the op-
position cobbled together a coalition government—the Alliance for European
Integration (AEI)—consisting of the four non-communist parties.
In turn, the AEI was in a much weaker position than the Communists to mo-
nopolize political control. As a fluid coalition of relatively equal political groups,
the AEI was extremely weak by the metrics used here. The parties were “partners
in a coalition” but also “political competitors” (Popescu 2012, 43). After taking
power, the AEI rapidly “decomposed into mutually hostile camps” (Socor 2013).
The regime incorporated strong “internal checks and balances” that made it
nearly impossible for any one faction to monopolize control (Popescu 2012, 43).
As in the 1990s, competing parties infiltrated different ministries and parliamen-
tary committees—which created an enormous amount of default competition
within the state and regime (Wilson 2013, 2).
As a result, Moldova opened up significantly. While deep antagonism to the
Communist Party allowed the coalition partners to unite in banning the pro-
Communist NIT TV, state-run media became more pluralist and the govern-
ment was no longer sufficiently unified to impose a single editorial line.32
Journalists no longer faced repeated harassment, and media felt free to air
programs critical of the government. Media coverage was balanced and reflected
a diversity of views in parliamentary elections in 2010, 2014, and 2019, as well
as in presidential elections in 2016 and 2020 (held after Moldova returned to a
semi-presidential system).33 In addition, the coalitional character of the govern-
ment made it nearly impossible for any single group to control the legislature.
Competition among coalition partners created a highly rapacious and competi-
tive political system in which leaders and parties repeatedly shifted alliances and
corruption was rampant. In the spring of 2013, Prime Minister Vlad Filat, from
the Liberal Democratic Party, came into open conflict with the Democrats, an-
other member of the coalition. In the end, Filat was forced to resign and was
replaced by Iurie Leancă in May 2013. The AEI was dissolved in favor of a new
“Pro European Coalition.” Between July 2009 and early 2021, the presidency was
controlled by five different politicians from different parties. During the period
of parliamentary rule that lasted until 2016, three presidents were elected by
the legislature: Mihai Ghimpu of the Liberal Party (2009–2010), Marian Lupu
of the Democrats (2010–12), and the independent Nicolae Timofti (2012–16).
In 2016, the country returned to semi-presidentialism and Igor Dodon of the
Socialists became president (2016–2020)—although Prime Minister Pavel Filip’s
Democratic Party of Moldova (PDM) was widely considered to be the ruling
party until 2019.34 In 2016, Dodon beat Maia Sandu of the pro-European Party of
Action and Solidarity and won the presidency in a close popular election. Then,
in 2020, Sandu defeated Dodon in free and fair elections. When this chapter was
written, Moldova could be considered a democracy.
Ukraine
Ukraine illustrates how party weakness and the persistence of old regime po-
litical institutions may foster democratic moments. Like Moldova, Ukraine
suffered from a weak civil society and has been riven by identity conflicts since
becoming independent in 1991. In the wake of the failed Soviet coup of August
1991, Leonid Kravchuk, a former chief of ideology in Soviet Ukraine, rode a wave
of anti-communism by destroying the Communist Party and pushing for the de-
struction of the Soviet state. Such moves allowed Kravchuk to survive the Soviet
collapse but meant that he came to power without any organized base of support.
Lacking a strong state and any political party, Kravchuk in 1991–94 was un-
able to control the legislature despite the fact that it was dominated by a nominal
ally, Ivan Pliushch. In 1992, Pliushch orchestrated the firing of a prime minister
loyal to Kravchuk, replacing him with Leonid Kuchma, the head of a missile fac-
tory in eastern Ukraine who had no ties to Kravchuk.35 After a wave of strikes in
eastern Ukraine in 1993, Kravchuk was forced to call early elections. Kravchuk
then changed his mind and decided to postpone elections and suppress parlia-
ment. However, he was forced to relent after heads of the police and security
forces refused to go along (Kravchuk 2002, 227–28). Simultaneously, Kravchuk’s
weak control over regional governments undermined efforts to manipulate the
election process, and Kravchuk’s own appointees often directly undermined the
president during the election and supported Kuchma (FBIS 1994; Kravchuk
2002, 230). As a result, Kravchuk lost power to Kuchma, who promised closer
ties to Russia.
The country descended into competitive authoritarianism under President
Leonid Kuchma (1994–2004). Under Kuchma, the government engaged in
systematic ballot stuffing, threats from government officials, and violence to
ensure victory. In the runup to the 2004 presidential elections, in which Kuchma
planned to install his chosen successor, Viktor Yanukovych, his regime poisoned
the main opposition leader, Viktor Yushchenko—permanently disfiguring his
face. However, efforts to steal the 2004 election sparked large-scale protests and
an electoral revolution that led to Yushchenko’s victory.
In 2005, Ukraine became a democracy. Under Yushchenko, the media was un-
constrained, and the opposition had total freedom to mobilize. Elections were
consistently free and fair and did not suffer from serious fraud (ODIHR 2006,
2007, 2010). In 2010, Yushchenko left office in a peaceful, democratic transition
of power to Viktor Yanukovych, who then put an end to Ukraine’s democratic
moment.
Given that the rise and fall of democracy directly coincided with Yushchenko’s
tenure, it is tempting to argue that Ukraine’s democratic moment was a product of
Yushchenko’s commitment to democracy. A closer look however, casts doubt on
this idea. When Yushchenko came to power, he immediately tried to attack the fi-
nancial bases of opposition forces—efforts that initially convinced Yanukovych’s
main backer to flee the country (Kudelia and Kuzio 2014). Yushchenko also fired
fourteen thousand civil servants for lack of loyalty to the new regime and used a
controversial interpretation of the constitution to shut down parliament and call
for early elections.36 It is thus difficult to claim that Yushchenko was committed
to pluralism.
While we can never know if even greater commitment to authoritarianism
might somehow have allowed Yushchenko to consolidate power, we do know that
he faced significantly greater challenges to his authority from within his own co-
alition than did either Kuchma before him or Yanukovych after him (see below).
Indeed, Yushchenko was backed by an extraordinarily weak coalition within
a coalition. “Our Ukraine” was a coalition of ten parties that ruled with Yulia
Tymoshenko’s Batkyvshina party. Before coming to power, Yushchenko created a
relatively equal alliance with Tymoshenko, who agreed in the summer of 2004 to
sit out the presidential election in exchange for her appointment as prime min-
ister. Competition between Tymoshenko and Yushchenko greatly hampered the
latter’s efforts to concentrate power, preventing him in particular from firing
judges and packing the courts with regime loyalists.37 As a result, Yushchenko
found it impossible to gain the necessary support in the Constitutional Court to
increase presidential powers (Trochev 2011). Divisions in the coalition also crit-
ically undermined Yushchenko’s capacity to use the state to manipulate elections
in his favor. Competition among allies for votes in the same areas served “as a
check on each other’s [electoral] shenanigans in the west and center.”38 As a re-
sult, electoral manipulation declined dramatically. In short, the coalitional struc-
ture of Yushchenko’s regime—in stark contrast to that of Kuchma—meant that
the president’s power was hemmed in from the start.
Political competition was also enabled by the persistence of old regime

institutions. Among the most important of these were media organizations with
ties to Yanukovych and his party. Even after his defeat and resignation in 2004,
pro-Yanukovych media ensured that he received positive or at least neutral cov-
erage (ODIHR 2006, 17–19). In general, although the Ukrainian press during
this period was wracked by endemic corruption, it was also quite pluralistic,
making it very hard for Yushchenko to impose a single official line. And with key
oligarchs supporting the opposition, a relatively equal playing field among the
major parties emerged.39
A key player in the opposition to Yushchenko was Yanukovych’s powerful
Party of Regions. Drawing on the support of extraordinarily wealthy energy
oligarchs and the country’s strongest regional political machine in Donetsk
province, the Party of Regions had “more money at its disposal” and was “more
organized” than other parties in Ukraine.40 It was also less personalized than
other Ukrainian parties, had the largest network of primary organizations in the
country, and retained strong voting discipline in the legislature.41 Yanukovych
was also backed by powerful oligarchs—including Viktor Medvedchuk and
Rinat Akhmetov, who controlled the major television stations 1 +1 and TV
Ukraina, respectively.
After Yanukovych was defeated in 2004, he did not disappear but
remained a powerful force. The Party of Regions possessed a large and loyal
base of electoral support in the east, and thus quickly became a key player
in the new Ukrainian political system. Yushchenko and Tymoshenko each
used Yanukovych to try to neutralize the other’s influence. In the summer
of 2005, Yushchenko obtained Yanukovych’s support to oust Tymoshenko
as prime minister in exchange for a promise to cease attacks on the latter
and provide amnesty for his 2004 electoral fraud.42 Yanukovych’s enormous
financial base therefore remained intact, providing his party with the re-
sources to capture a plurality of seats in the 2006 parliamentary elections
and ultimately seize the prime minister’s office (after Yushchenko and
Tymoshenko failed to forge a coalition).43 In the 2010 presidential elections,
both Tymoshenko and Yushchenko ran against Yanukovych. Yushchenko—
whose campaign was wracked by defections, severe underfunding, and
disorganization—obtained just 5 percent of the vote.44 Tymoshenko made it
to the second round against Yanukovych, who beat Tymoshenko 49 percent
to 45 percent in free and fair elections.
As in Moldova, the emergence of a better organized party put an end to
Ukraine’s democratic moment. Drawing on a disciplined party and a compliant
prime minister, Yanukovych packed the courts and increased the formal powers
of the president. Under the guise of combating corruption, the government jailed
Tymoshenko and other opposition activists.
However, a third democratic moment emerged after Yanukovych fell in 2014

to large-scale protests that both forced the president to flee the country and led
to civil war. Large sections of the Ukrainian state collapsed. Police and special
forces “disappeared from the streets” (Puglisi 2014). The new authorities in Kyiv
lacked any effective means of imposing order—outsourcing security functions to
local oligarchs and opposition paramilitaries (Puglisi 2014). Ukrainian president
Petro Poroshenko came to power in 2014 with a party that was “no more than a
myth,” with no website or known address or telephone number.45
War with Russia created challenging circumstances for democratic develop-
ment. In significant parts of eastern Ukraine, weakened security made it impos-
sible to carry out the election. Furthermore, the conflict threatened democracy
by encouraging government efforts to restrict opposition Russophile media. For
example, in 2018, the opposition Radio Vesti, operating in Kyiv and Kharkiv, was
shut down due to its “rude and derogatory remarks addressed to the heroes [of
the EuroMaidan Revolution]” (quoted in Way 2019, 55). Therefore, Ukraine
cannot be considered a full democracy between 2014 and 2019.
Nevertheless, the opposition was given substantially greater freedom to op-
erate than under Yanukovych, and politics were extraordinarily open and com-
petitive in those parts of the country under Ukrainian control. In parliamentary
and presidential elections in 2014 and 2019, there was little fraud and the media,
while biased, represented diverse and opposing political forces.46 Each election
witnessed dramatic shifts in political power. Most notably, in 2019, Poroshenko
was defeated 25 percent to 73 percent in presidential elections while his party
lost 80 percent of its seats in parliament.
Democratic success may partly be traced to Ukraine’s vulnerable interna-
tional position. Under attack from a power three times its size, the government
was heavily dependent on Western military and economic support. As a result,
Poroshenko faced constraints in undertaking overtly undemocratic behavior
that might undermine the country’s international reputation. Such constraints
may have prevented democratic backsliding early in Poroshenko’s term. Thus, it
was rumored that President Obama successfully discouraged Poroshenko from
imposing martial law when the two leaders met in June 2014. Perhaps more im-
portantly, Poroshenko was unable to convince his own allies to support limita-
tions on democracy. In late 2018, following a Russian naval attack, Poroshenko
did declare martial law, which would have postponed upcoming presidential
elections. However, this move sparked widespread opposition from within
his own supporters and the Ukrainian elite in general. As a result, Poroshenko
amended the law and held the elections as scheduled. He was then soundly de-
feated by a total outsider, the comedian Volodymyr Zelensky. Under Zelensky,
Ukraine has maintained similar levels of democracy: restrictions on Russophile
media (supported by the Biden administration)47 have continued at the same
time that the system has remained extremely open and competitive. In 2021,
Ukraine was very nearly democratic.
Georgia
Suffering a complete breakdown of social order and descent into warlord

rule in the early 1990s, Georgia arguably faced more serious obstacles to de-
mocracy than Moldova and Ukraine.48 Perhaps as a result, Georgia since in-
dependence has not been as democratic as Moldova or Ukraine.49 Georgia
nevertheless experienced a surprising democratic moment in the 2010s. In
2012, the country witnessed the first peaceful democratic turnover of power
in its history. In subsequent years, elections improved in quality and the oppo-
sition faced relatively few constraints in mobilizing support before early 2021,
when the government arrested the main opposition leader. While short-lived,
Georgia’s democratic moment is unexpected given the country’s unpromising
conditions. This outcome, I argue, can best be explained by Georgia’s vulner-
ability to external pressure, the weakness of the ruling party, and the strength
of old regime forces that remained intact after the fall of President Mikheil
Saakashvili in 2013.
After becoming independent in 1991, Georgia experienced a short-lived dic-
tatorship under Zviad Gamsakhurdia that led to total state breakdown and civil
war in 1991–92, and ethnic conflict involving the country’s ethnic enclaves of
Ossetia and Abkhazia, as well as Russian military interventions in the early 1990s
and in 2008. While no dictator lasted for long, Georgia also lacked a strong op-
position. Opposition challenges have been weakened by the fact that old regime
forces tended to disappear completely following their ouster—leaving the new
government with fewer checks on its power. In 1991, 1992, and 2004, new leaders
came to power with more than 80 percent support before completely collapsing
as political forces. Each time, the near total collapse of the prior regime (the
Communists in 1991, Gamsakhurdia in 1992, and Shevardnadze in 2003) meant
that new leaders ruled without an effective political opposition that could check
their power.
Georgia’s democratic moment emerged out of the Rose Revolution of 2003.
In the late 1990s and early 2000s, Georgia was ruled by President Eduard
Shevardnadze, who had been brought in by a warlord in the early 1990s in the
wake of the collapse of the Gamsakhurdia regime. While Shevardnadze brought
a measure of stability, corruption was rampant, the central government faced
regional challenges in Ajaria, Ossetia, and Abkhazia, and the police often
went unpaid for months at a time (see Devdariani 2003). As the popularity of
Shevardnadze declined in the early 2000s, Shevardnadze’s government fell apart.
Top officials—including most notably Justice Minister Mikheil Saakashvili—

split from the president and went into opposition.
Marked by significant fraud, parliamentary elections in 2003 sparked mass
protests after official results showed Shevardnadze’s coalition winning a narrow
plurality. Confronting an underfunded police force, Saakashvili easily stormed
parliament and forced Shevardnadze to flee.50 At the same time, Shevardnadze
was isolated internationally. He lacked support from Russia and was highly de-
pendent on US aid.51 In mid-2003, President Bush sent Shevardnadze a letter
expressing hope that he would cede power to a “new generation of leaders”
(Devdariani 2004). Unable to rely on the weakened coercive apparatus and iso-
lated from both Russia and the West, Shevardnadze had little choice but to resign.
Like previous Georgian governments, Shevardnadze disappeared as a political
force. Power shifted dramatically to Mikheil Saakashvili and his United National
Movement (UNM). Saakashvili won the presidency with 96 percent of the vote
and his UNM captured nearly two-thirds of parliament. Saakashvili quickly
built a relatively well-disciplined party and strengthened the state. Reforms
by Saakashvili increased the coercive capacity of the police, and key regions—
such as Ajaria, a province on the Black Sea coast that had enjoyed de facto au-
tonomy from the weak central government—were brought under central control
(Fairbanks and Gugushvili 2013, 117).
Saakashvili’s government was not democratic.52 Media harassment persisted,
including tax raids of independent television stations, prosecution of journalists,
and government pressure to cancel programs critical of Saakashvili (see Dolidze
2007; Fuller 2005; Peuch 2004). In 2008, Saakashvili was reelected in a presiden-
tial election that was marred by harassment of opposition supporters and nu-
merous irregularities in voting (ODIHR 2008).
Contrary to widespread expectations, however, Saakashvili’s tenure ended in
an important democratic moment made possible by Western pressure, incum-
bent weakness, and the survival of old regime forces. First, Western involve-
ment in Georgian politics created future opportunities for international actors
to discourage autocratic measures. There existed strong public and elite support
for the European Union in Georgia (Muller 2011). Furthermore, in the wake of
the Russian invasion of 2008, the country was heavily dependent on Western
military assistance—relying on the United States for thirty percent of its mili-
tary expenditures.53 Georgia’s increasing dependence on the West as a bulwark
against Russian aggression made it potentially vulnerable to democratizing
pressure.
Opposition to Saakashvili began to coalesce around a powerful Georgian
billionaire, Bidzina Ivanishvili, after the 2008 election. Educated in Moscow in
the 1980s, Ivanishvili had amassed significant wealth in the privatization gold
rush in Russia in the early 1990s. In 2013, Forbes estimated that his wealth was
$ 5.3 billion—almost half of Georgia’s annual GDP (Gente 2013, 4). A Georgian
citizen, Ivanishvili invested heavily in Georgia—paying to repair Tbilisi State
University and building hospitals and parks (van Peski 2013, 72). He initially
backed Saakashvili (Gente 2013, 6), but his support shifted after the 2008 presi-
dential election (Gente 2013, 7). Criticizing government abuse, Ivanishvili began
funding opposition parties (Gente 2013, 7). Then, less than a year before the 2012
parliamentary elections, he created the “Georgian Dream” coalition consisting of
six opposition parties that included a mix of NGO activists, a chess champion,
and a football player (Aprasidze 2013, 223; Fairbanks and Gugushvili 2013, 119;
van Peski 2013). The coalition lacked any ideological coherence—combining
European-minded liberals and hardcore Georgian nationalists (Gilbreath 2015,
107; van Peski 2013, 72). At the same time, Ivanishvili’s vast riches meant that the
coalition was much more centralized than coalitions described above in Ukraine
under Yushchenko and Moldova under the AEI.
Saakashvili responded aggressively to Ivanishvili’s decision to go into op-
position. Accused of being a “Kremlin stooge,” Ivanishvili was deprived of his
Georgian citizenship shortly after he announced his intention to create the
Georgian Dream coalition.54 The government began harassing Ivanishvili’s
supporters, and imposed enormous fines on Ivanishvili and his allies (Cecire
2013, 237; Fairbanks and Gugushvili 2013, 120, 121; Freedom House 2014, 228).
However, two factors hampered Saakashvili’s efforts to quash opposition. First,
the opposition was powerful and united. In contrast to earlier oligarchs who op-
posed Saakashvili, Ivanishvili did a much better job coopting and unifying oppo-
sition parties—allowing Ivanishvili to present a united front that was backed by
an extraordinary amount of money. Second, in part due to the strength of the op-
position, it was harder for Saakashvili to garner Western support. In the absence
of any serious challenges to Saakashvili in the early 2000s, unconditional support
had been costless to the West. However, in the runup to the 2012 parliamen-
tary election, President Barack Obama faced the prospect of backing Saakashvili
amid protests provoked in reaction to a stolen election.55 Thus, when Obama
met with Saakashvili in early February 2012, Obama stressed that the conduct
of the upcoming elections would be a “litmus test” for US–Georgian relations.56
Simultaneously, various international agencies funded by the US government
gave extensive coverage of Saakashvili’s abuses (Cecire 2013, 247; Mueller 2014,
344). Finally, Ivanishvili engaged significantly with the United States. He met
regularly with the US ambassador to Georgia and hired a number of Western
lobbyists to bolster his image in the West.57 In early 2012, Ivanishvili paid to take
out a full-page ad in both the New York Times and the Washington Post making
the case for free and fair elections in his country.58 Saakashvili, sensitive to inter-
national criticism, passed a “must carry” law that required all cable services to
carry a diversity of news channels in order to ensure the population access to all
types of information (Freedom House 2013).
In the end, the Georgian government backed off from attempts to steal the
election from Ivanishvili. Saakashvili did not yield on the citizenship question
but instead made a change in the constitution to allow Ivanishvili as a European
Union citizen to run in Georgian elections.59 Though in August the govern-
ment declared that Ivanishvili’s coalition would be barred from taking part in
the elections due to violations of spending laws, it “flinched” in the end, fearing
foreign disapproval (Fairbanks and Gugushvili 2013, 121). As a result, the cam-
paign was relatively free and fair. The media gave Georgian Dream significant
coverage (ODIHR 2012, 2). While Saakashvili’s government had tried to pad the
electoral rolls in the hopes of manipulating the vote, the margin of victory for
Georgian Dream was simply too large and efforts at manipulation were seem-
ingly abandoned (Fairbanks and Gugushvili 2013, 123; Mitchell 2013, 79–80).
The Georgian Dream coalition obtained 85 seats and the National Movement 65.
Faced with such a clear outcome and internationally isolated, Saakashvili took
to the airwaves to concede his party’s defeat and announce that it would go into
opposition (Cecire 2013, 239; Fairbanks and Gugushvili 2013, 123). Despite re-
taining control of the presidency, he gave up entire control of the government to
Georgian Dream (Kvashilava 2019, 230). (Georgia was a semi-presidential re-
gime from 2004 to 2018.)
As the first peaceful, democratic turnover in Georgia’s history, the 2012
election marked a major milestone in Georgia’s democratic development. The
country became “unquestionably the most open polity of the South Caucasus”
(Cornell 2014, 185; see also Fumagalli 2014, 396) and more democratic than any
other post-Soviet state (van Peski 2013, 50). In the words of one commentator,
politics in Georgia became “boring” (Gilbreath 2015).
The dramatic transformation of Georgian politics during this period can be
seen in the improvement of the quality of elections in 2012, 2013, 2016, and 2018
(Mitchell 2016; see also Freedom House 2013, 2015, 2019b). Each election was
highly competitive.60 In contrast to Saakashvili, the Georgian Dream govern-
ment during this period relied less on coercion and much more on cooptation to
remain in power (Freedom House 2019b). The political system no longer saw the
widespread use of government-sponsored threats and violence that had plagued
Georgian politics in its first two decades of independence. In contrast to elections
in the past, the opposition was given the freedom to campaign in all parts of
the territory under government control. Incidents of violence became “isolated”
(ODIHR 2017, 2). In elections in 2013 and 2016, international monitors carrying
out parallel vote counts confirmed the accuracy of the official results (NDI 2013;
NDI 2016, 4). Many issues raised by international actors—including the use of
government web pages for campaigning—were ones that exist in countries that
most consider to be democratic (ODIHR 2019, 11–12).
As electoral quality improved under Georgian Dream, so too did media plu-
ralism. While some popular talk shows were unexpectedly canceled and govern-
ment leaders expressed frustration that media were “circulating lies and creating
tension in the society” (Freedom House 2015, 2016), media were generally vi-
brant and frequently aired critical voices (Freedom House 2017; Mitchell 2016).
Georgian television frequently featured lively debates and a diverse range of views
(Freedom House 2015). Candidates during election time had nearly unfettered
access to media. Due in part to laws passed under Saakashvili in 2012, media
coverage during election was relatively balanced (ODIHR 2016, 19). Finally, civil
society operated without harassment and nongovernmental organizations were
given ample opportunity to express critical opinions (Freedom House 2015).
The surprising emergence of genuine, if flawed, pluralism can be at least partly
explained by the persistence of old regime forces and party weakness. First, the
new regime was heavily shaped by the persistence of Saakashvili-era media
and Saakashvili’s party. As noted above, previous transitions had witnessed
the complete disintegration of the old government—leaving the new adminis-
tration without serious competitors. Things were different in 2012. Prominent
pro-Saakashvili forces—the UNM and the pro-Saakashvili Rustavi 2 television
station—remained in place after he left power, creating important checks on the
new government. In particular, the United National Movement remained a po-
tent political force—even after Saakashvili left the country in 2015. The UNM
had an ideology focusing on reform and a core group of convinced loyalists in a
position to challenge the new administration (Fairbanks and Gugushvili 2013,
124; Orovec and Holland 2019, 39). In 2012, the party won 40 percent of the
vote, and in 2018, it won a plurality and forced the Presidential election into two
rounds—before being defeated by the Georgian Dream candidate.
Most importantly, Rustavi 2, the most popular television station in Georgia,
watched by 80 percent of the public, provided an important democratic check
on Georgian Dream. A critical source of pro- opposition coverage under
Shevardnadze, the station had been transformed from a government critic into
a pro-Saakashvili “propaganda machine” after 2003.61 But, under the director-
ship of Nika Gvarmia, a former UNM deputy and minister of justice, it survived
Saakashvili’s fall. The station aired hundreds of stories exposing government cor-
ruption and Ivanishvili’s business activities.62 In the 2013, 2016, and 2018 na-
tional elections, it provided a prominent source of criticism of the incumbent.
Rustavi 2’s persistence was not a foregone conclusion. Indeed, immediately
after taking power, the government began an effort to silence the station by
transferring ownership to Kibar Khalvashi, a government ally—an avenue made
possible by questionable changes of ownership under Saakashvili (Welt 2015). In
mid-2015, a Tbilisi court supported the transfer of ownership to Khalvashi, who

claimed that his shares had been illegally liquidated by Saakashvili’s government
in 2005. However, international outcry convinced the government to allow the
station to remain in the hands of pro-UNM forces until the case was ruled on by
the European Court of Human Rights. As a result, the station continued to air
criticism of the government (Freedom House 2017).
Second, Georgian pluralism was bolstered by divisions within Georgian
Dream itself. Founded as a loose coalition of ideologically diverse forces just a
few months before it came to power, the party suffered significant defections—
especially in its first years in government (see Aivazian 2019, 112–13). These
divisions undermined government efforts to curb the opposition. In 2013,
Giorgi Margvelashvili, a former minister of education in the Georgian Dream
government and a member of the opposition under Shevardnadze (Fumagalli
2014, 396), won presidential elections with 62 percent of the vote. After as-
suming office, Margvelashvili became a check on the prime minister’s power
and “a clear voice for democracy, reform and human rights in Georgia” despite
being a member of the same party (Freedom House 2016, 2017; Mitchell 2016).
Together with the Georgian Dream parliamentary chairman, Margvelashvili op-
posed government attacks on Rustavi 2 (Bukia 2015; Welt 2015). Margvelashvili
successfully thwarted efforts by the government to weaken bureaucratic inde-
pendence and preserve government surveillance capacities inherited from
Saakashvili (Freedom House 2016). Similarly, the Georgian Dream speaker
of Parliament, Davit Usupashvili, condemned police attacks on protesters
(Freedom House 2014).
At the same time, problems existed. After taking power, the government reg-
ularly used state agencies to target potential sources of opposition. Fulfilling
a campaign promise, Ivanishvili immediately began to target UNM officials
for prosecution (Cecire 2013, 237; Fumagalli 2014, 397; Human Rights Watch
2015). By the end of 2012, more than twenty former government officials had
been detained on criminal charges (Freedom House 2013, 2014). In 2014, the
government began efforts to prosecute Saakashvili himself for abuse of office.
The government seized his property and tried but failed to convince Interpol to
put Saakashvili on its international wanted list.63 In response, Saakashvili was
forced to flee the country. Media independence, too, suffered. In July 2019, the
European Court for Human Rights found no evidence of government interfer-
ence in Georgian court decisions transferring ownership of the opposition tele-
vision station Rustavi 2 to a Georgian Dream ally.64
Increased problems might be linked to the fact that the Georgian Dream coa-
lition became less fractured over time as members of the coalition defected from
the core group controlled by Ivanishvili. In 2016, Georgian Dream won a ma-
jority 115 of 150 seats in the legislature. In the runup to the 2020 parliamentary
elections, the government began carrying out more serious attacks on the oppo-
sition. Shortly after the banker Mamuka Khazaradze announced his intention
to form an opposition bloc to challenge Georgian Dream, prosecutors brought
fraud charges for a violation that had allegedly occurred eleven years prior65—a
move that was widely interpreted as an open attack on the opposition.66 After
parliamentary elections in late 2020, opposition parties alleged fraud and
boycotted legislative sessions.67 Then, in early 2021, police raided the headquar-
ters of the UNM and arrested the party’s leader, Nika Melia.
At the same time, electoral life remains more pluralistic than in Georgia’s past
or most of the rest of the former Soviet Union. Indeed, the 2020 parliamentary
elections were characterized by open campaigning of the opposition and wide
coverage of the major parties. Opposition leader Khazaredze’s party was allowed
to campaign for the parliamentary election and gained seats in the legislature.68
The kind of pervasive violence that characterized Georgia’s first decades of inde-
pendence was almost entirely absent. Finally, in 2021 there was still some hope
that internal divisions might prevent a more complete descent into authoritari-
anism. Even though the party was much more unified than it had been in 2013,
it remained divided over the regime’s authoritarian trajectory. Thus, the arrest
of Melia caused a major rupture within Georgian Dream. The Prime Minister
Giorgi Gakharia resigned in protest over Melia’s arrest.
Conclusion
In contrast to other chapters in this volume, Georgia, Moldova, and Ukraine did
not witness the emergence of stable democracy. At the same time, an exclusive
focus on the global failure of democracy in the post-Soviet region blinds us to
very real instances of democratic success that emerged even in hostile conditions.
During the “democratic moments” described in this chapter, elections were (very
nearly) free and fair, the opposition was given free reign, and media gave exten-
sive and often positive coverage to those who were critical of the government. In
such cases, incumbents confronted intense political competition and often lost
power in peaceful and democratic elections. At the same time, democracy was
not consolidated in the sense that leaders regularly tried to abuse their power
and change the fundamental rules of the game. Such moments were particularly
evident in Georgia, Moldova, and Ukraine—that had on average the greatest
democratic success in the former Soviet Union since 1991. The chapter focuses
on three such moments in Ukraine, two in Moldova, and one in Georgia.
I argue that three factors helped facilitate democratic moments in the cases
above. First, authoritarian weakness—in particular the weakness of authori-
tarian states and ruling parties—undermined efforts to consolidate authoritarian
control. Such weakness hardly guaranteed democracy, but it created the possi-
bility of pluralism in the absence of a robust civil society or well-organized op-
position. In each of these cases, weak parties and weak states undermined efforts
to suppress opposition challenges and contributed to relatively free and highly
competitive elections. Second, the persistence of parties and media from the
previous authoritarian government created key checks on leaders and enhanced
the likelihood of a democratic moment. In both Ukraine in the late 2000s and
Georgia in the 2010s, old regime media and parties provided an important
source of democratic contestation that hemmed in the autocratic ambitions of
incumbents (Loxton and Mainwaring 2018). Third, and finally, Russian inter-
vention in the region had a contradictory impact. On one hand, it likely strength-
ened pro-Russian authoritarian forces such as Kuchma and the Party of Regions
in Ukraine and the Communist Party in Moldova. It has also periodically moti-
vated anti-Russian governments to suppress opposition Russophile media. At
the same, it has left anti-Russian governments vulnerable to Western democra-
tizing pressure. Such pressure clearly played a role in convincing Saakashvili to
step down in Georgia in 2012–13 and may have undermined efforts to consoli-
date authoritarian rule in Ukraine in 2014.
What does this analysis add to the discussion of democracy in hard places?
First, it suggests that important cases of democratic success can emerge short
of long-term consolidation. (Indeed, the breakdown of democracy in Benin in
2019 and the democratic crisis in India show that cases that might seem con-
solidated are often not so.) The democratic moments analyzed in this chapter
were relatively short-lived but quite real and thus require explanation. Second,
in line with the analyses of Reidl and Slater (Chapters 4 and 3, respectively, this
volume) and contra Varshney and Mainwaring (Chapters 2 and 7, respectively,
this volume), this chapter suggests that democracy emerges less from normative
commitment and more from constraints on behavior. While we can never know
for sure if leaders might somehow have monopolized power if they had been more
autocratically inclined, we do know that each leader willingly engaged in anti-
democratic behavior and faced identifiable constraints on authoritarian behavior.
Finally, the argument presented here may account for democracy in other
hard places. In particular, pluralism by default appears to explain the emergence
of democratic competition in some parts of sub-Saharan Africa, which, like the
former Soviet Union, includes numerous countries with relatively weak states
and ruling parties. For example, incumbent weakness at least partly explains
Benin’s long and very surprising period of democratic success that is analyzed
in Chapter 4, by Rachel Reidl. (For other examples of pluralism by default in the
African context, see Levitsky and Way 2010, chap. 6.)69 Comparing Benin and
South Africa, Reidl argues that democracy in Benin emerged out of a balance
of forces between the incumbent and the opposition. Yet, given the weakness of
both the opposition and civil society in Benin, it is not obvious how such a bal-
ance emerged after the country’s transition to democracy in 1991. (In econom-
ically developed South Africa, which had a robust civil society, such a balance
seems less surprising.) Indeed, opponents of Mathieu Kérékou’s military dicta-
torship in Benin were quite weak in the late 1980s and early 1990s (Allen 1992,
74; Nwajiaku 1994, 431; Riedl 2014, 161).70
A closer look at Benin suggests that pluralism by default may explain the bal-
ance of forces described by Reidl. Here, we see interesting parallels to the post-
Soviet examples of pluralism by default described above. As in many post-Soviet
cases, opposition to the old regime in Benin emerged less from a robust civil so-
ciety and much more from the fragmentation of the old regime.71 Furthermore,
the ruling Peoples’ Revolutionary Party of Benin (PRPB), like the Soviet
Communist Party, completely disappeared during the transition. In April 1991,
following a National Conference in which PRPB officials called for an end to dic-
tatorship, the party effectively dissolved itself (Allen 1992, 72; Riedl 2014, 166).
Like post-communist apparatchiks who rapidly abandoned the Communist
Party in 1991 to save their own skins, former supporters of the PRPB felt that
dissolving the party was their best root to survival (Riedl 2014, 166). In turn, the
rapid demise of the PRPB resulted in a political vacuum and the proliferation of
an extremely large number of very weak parties with particularistic and narrow
geographic bases of power (Kuenzi and Lambright 2001, 450; Riedl 2014, 197–
98). Finally, as in many parts of the former Soviet Union, Benin emerged from
the transition with a weak state (Heilbrunn 1993, 285).72
The combination of party and state weakness created prime conditions for
pluralism by default. Between 1991 and 2016, democratic competition in Benin
can be directly traced to party and state weakness. First, in the March 1991
elections that pitted Kérékou against Soglo, President Kérékou was deprived of
virtually any incumbent advantages. Kérékou ran without a party. The opposi-
tion leader, Nicéphore Soglo, who had been appointed as prime minister by the
National Conference, was given control over the military. All of Kérékou’s allies
in the government were removed (Englebert 1996, 164–65). Soglo trounced
Kérékou in the second round 68 percent to 32 percent.
Soglo, in turn, came to power with “neither a political base nor a political fol-
lowing” (Amuwo 2003, 163).73 He was forced to “cobble together” a three-party
coalition that collapsed within a year (Decalo 1997, 57; Englebert 1996, 166).
Because of Soglo’s inability to create a stable majority, the legislature gained sig-
nificant de facto power and voted to establish an autonomous electoral commis-
sion that was opposed by Soglo (Englebert 1996, 166; Magnusson 1999, 225).
Soglo also lacked firm control over the means of coercion.74 Both opposition and
incumbent had paramilitaries with the capacity for intimidation (Amuwo 2003,
166). As a result, Soglo enjoyed relatively few incumbent advantages in the runup
to the 1996 election, when Soglo again faced Kérékou in the second round. After
losing to Kérékou by five points, Soglo—the incumbent—“alleged massive rig-
ging” and threatened to contest the results in the courts. However, facing little
chance of success, he backed down and accepted defeat (Amuwo 2003, 166–67;
Levitsky and Way 2010, 295).
While Kérékou lacked a strong party,75he was a relatively skilled backroom
politician and pulled together a coalition of elite supporters—buying the support
of Houngbédji by naming him prime minister (Levitsky and Way 2010, 295). In
2001, Kérékou won reelection. However, Kérékou failed to keep the support of
Houngbédji, whose Parti du Renouveau Démocratique (PRD) became one of the
main opposition parties in the country (Africa Confidential 2002). Without a
party, Kérékou was unable in 2006 to decide on a successor—creating an election
that was wide open (Levitsky and Way 2010, 296). While Kérékou made a con-
certed effort to cancel elections and hold onto power, domestic and international
backlash convinced him to let the elections go forward (Africa Confidential
2006; Seely 2007, 198). As a result, Boni Yayi, a political outsider, won without the
backing of a party in an election that was widely considered free and fair (Africa
Confidential 2006; Seely 2007, 197–99). After winning reelection in 2011, Yayi
sought to change the constitution and seek a third term.76 He was only dissuaded
after a series of defections from his party, Cowry Forces for an Emerging Benin,
and this party’s failure to secure a majority in the 2015 legislative elections.77 In
2016, Patrice Talon, a former financial backer of Yayi who had fallen out with the
president, beat out Yayi’s chosen successor in relatively free elections.78
The dynamics described above suggest that, to an important extent, the bal-
ance of forces in Benin were rooted in incumbent weakness. Such weakness both
thwarted efforts by leaders to engage in backsliding and allowed relatively unor-
ganized opposition (all of whom ran as independents) to successfully challenge
incumbent power.
However, pluralism by default is not a durable basis for democracy and thus
we would not expect pluralism to survive indefinitely. Indeed, after 2016, Benin
slipped unambiguously back into competitive authoritarian rule. Sébastien
Ajavon, a successful businessman who came in third in the 2016 election, was
forced into exile after being sentenced in absentia for a drug trafficking charge
that many considered to be politically motivated.79 The legislature also voted in
July 2018 to lift the immunity of three opposition deputies so that they could
face corruption charges.80 Then, the government passed a new electoral law in
September that effectively made it impossible for opposition parties to compete
in the 2019 parliamentary elections.81 As a result, the National Assembly only in-
cluded government supporters.
In sum, there is at least some evidence of pluralism by default outside the
former Soviet Union. As in Georgia, Moldova, and Ukraine, state and party
weakness undermined efforts by successive incumbents to consolidate political

power in Benin. And, as in the post-Soviet cases, democracy never consolidated
in Benin. Leaders repeatedly sought to undermine the rules of the game, and de-
mocracy collapsed by the late 2010s. At the same time, the comparison between
post-Soviet cases, on one hand, and both Benin and the other cases covered in
this book, on the other, raises an important question. Why did democracy in the
former Soviet Union fare worse than in the cases covered in the book? Indeed,
democracy survived for at least twenty-five years in Benin—and longer in the
other cases covered in the book. What might have prevented post-Soviet cases
from establishing more stable democratic rule?
One possible answer lies in the differing regional environments. In contrast
to the other cases covered in this book, post-Soviet cases have been shaped by
the existence of a dominant regional autocracy intent on interfering with the
politics of smaller countries. As discussed above, Russia has influenced re-
gime politics in contradictory ways. On one hand, Russian threats have pe-
riodically enhanced these countries’ vulnerability to Western democratizing
pressures and strengthened challenges to their rule. At the same time, such in-
terference has undermined democracy in two ways. First, as we saw in Ukraine,
Russian security threats encouraged the government to suppress opposition
Russophile media. Furthermore, the Russian government has given support to
pro-Russian autocrats in Moldova and Ukraine. Thus, in both countries, dem-
ocratic moments were cut short by the election of pro-Russian politicians and
parties (President Kuchma and the Party of Regions in Ukraine; the Communist
Party in Moldova). While the case studies presented evidence that organizational
factors helped undermine democracy, it is also possible that Russian influence
pushed these governments toward greater autocracy as well. Russian influence
is too contradictory and/or weak to stamp out pluralism, but, given the already
weak conditions for democracy in the region, it may make true democracy in
hard places harder to maintain for significant periods of time.
Notes
1. In line with other scholars of the region, I use “post-Soviet” to describe the twelve re-
publics of the Soviet Union excluding the Baltic states, which entered the Soviet Union
much later than the other republics and had traditions of independent statehood the
others lacked.
2. “Consolidated” democracies are democratic regimes in which major rules regulating
the structure and acquisition of power change very infrequently and are taken for
granted by virtually all major regime actors.
3. See, for example, Anderson 1999; Fish 1995; Hough 1997; and McFaul 1999; for more
pessimistic early assessments, see Jowitt 1992 and Roeder 1994.
4. For example, Levon Ter-Petrosian in Armenia, Yeltsin in Russia, and Kuchma in

Ukraine all continued to receive significant aid following serious democratic abuses
in the mid-and late 1990s. See Bureau of European and Eurasian Affairs 2009.
5. Such peaceful democratic turnovers took place in Belarus in 1994; Moldova in 1996,
2001, and 2019; Ukraine in 1994, 2010, and 2019; and Georgia in 2013.
6. Compared to my own assessment, V-Dem is rather easy on Georgia after 2003
and puzzlingly stringent on Ukraine after 2014, showing a decline in democracy
after the Euromaidan revolution of 2014. Unfortunately, V-Dem’s reliance on ex-
pert surveys does not allow us to identify which specific incidents or government
behavior (or whether such incidents/government behavior took place in parts of
the country controlled by the national government) motivated V-Dem’s divergent
assessments.
7. For example, Russia’s President Boris Yeltsin was clearly more tolerant of media criti-
cism than his successor, Vladimir Putin.
8. The coding of post-Soviet cases is based on World Bank data on natural resource
wealth as share of exports (World Development Indicators, n.d.).
9. The other nonresource cases varied between those that were highly authoritarian
(Belarus and Tajikistan) and competitive authoritarian (Armenia and Kyrgyzstan).
10. Incumbents backed by weak ruling parties are those who are supported by no party,
who are part of a ruling coalition of multiple parties, or who are running backed by a
new (and therefore not well established) party that has not run in two elections. See
Way (2015) for a full discussion of causal variables and codings.
11. “Nationalists Send Party to Scrapyard as They Take Control,” The Times (London),
August 31, 1991.
12. This concept is similar to Michael Mann’s (1984) “despotic power” and Brian Taylor’s
(2011) “exceptional state power.”
13. As Sidney Tarrow notes, “[r]‌ ational people do not often attack well- fortified
opponents” (Tarrow 1996, 54). See also Beissinger 2002, 152–53.
14. Helsinki Watch 1993, 18– 19; Waters 1996, 398; author interviews with Viorel
Cibatoru (former military advisor), Chişinău, Moldova, February 7, 2002; and
Nicolai Chirtoaca (former National Security advisor), February 5, 2002.
15. It is coded as weak according to my criteria because when it first backed Snegur, it had
not run in any elections.
16. Author interview with Alexandru Muravschi (parliamentary deputy from 1994 to
2001), Chişinău, January 31, 2002.
17. Ibid.
18. “Presidential Campaign in Moldova Is Likely to Have Its Own ‘Lebed,’ ” Infotag,
July 13, 1996.
19. Author interview with Adrian Usatii, Chişinău, July 30, 2004.
20. “Moldova: Parliament Speaker Calls for Radio, TV Chief ’s Ouster,” Infotag, July 28, 1997.
21. In the elections, the Communists gained forty seats; the BDPM, twenty-four; the
Democratic Convention of Moldova, twenty-six; and the Party of Democratic Forces,
eleven.
22. Author interview with Dumitru Diacov (head of parliament, 1998–2000), Chişinău,
February 1, 2002.
23. Ibid.; author interview with Alexandru Mosanu (leader of the Front), Chişinău, February
4, 2002; author interview with Anatol Golea (journalist), Chişinău, February 1, 2002.
24. Diacov interview, 2002.
25. Quoted in “Political Commentary Examines Ousted TV Chief ’s Case,” Basapress, July
30, 1997. See also Diacov interview, 2002; Golea interview, 2002; Muravschi inter-
view, 2002.
26. “Moldova: Parliament Approves New Teleradio Management,” Infotag, November
30, 1997.
27. Author interview with Angela Sarbu, Independent Journalism Center. Chişinău,
February 6, 2002.
28. For details of these proposed amendments, see Roper 2001, 12.
29. RFE/RL Newsline, December 16, 1999.
30. Diacov interview, 2002.
31. For a detailed description, see Way 2015, chap. 3.
32. Because of this closure, Moldova cannot be considered a full democracy during this
period.
33. See OSCE election reports for these elections in “Elections in Moldova,” at https://
www.osce.org/odihr/elections/moldova/.
34. See Robert Petraru, “Is Moldova Heading into Uncertainty as the Democratic Party
of Moldova Grows Stronger?,” Vocaleurope, March 4, 2019, https://www.vocaleurope.
eu/is-moldova-heading-into-uncertainty-as-the-democratic-party-of-moldova-
grows-stronger/.
35. Ukraine is a semi-presidential system. While the formal powers of the president have
fluctuated since 1991, the president has consistently been considered the most pow-
erful figure in the country.
36. See Trochev (2010) and “Президент и правящая Партия регионов не только
постарались укрепить вертикаль. . .” [The President and the ruling Party of
Regions not only tried to strengthen the vertical of executive power . . .], Зеркало
Недели, September 25, 2010.
37. Trochev (2010, 136); “Голубая контрреволюция: Виктор Янукович ликвидирует
последствия ‘оранжевой революцииь’ ” [Blue counterrevolution: Viktor
Yanukovych liquidates the consequences of the ‘Orange Revolution’], Gazeta.Ru,
September 30, 2010.
38. US Embassy Cable, September 7, 2007.
39. James Marson and Dariya Orlova, “Nation’s News Media: Free or Still Captive to
Their Owners?,” Kyiv Post, October 15, 2008.
40. US Embassy Cable September 7, 2007. See also Kushnarev 2007.
41. US Embassy Cable, September 28, 2007; “ ‘Партия регионов и “Единый центр’
имеют самые большие сети первичных организаций в стране –Минюст” [The
Party of Regions and the ‘United Center’ have the largest networks of primary organ-
izations in the country—Ministry of Justice], Униан, March 24, 2010; US Embassy
Cable, April 15, 2008.
42. “Меморандум порозуміння між владою та опозицією” [Agreement memo-
randum between the government and opposition], Ukrainska Pravda, September 22,
2005. Quoted in Kudelia and Kuzio 2014.
43. The Orange coalition (including Moroz’s Socialist Party) garnered just over 40 percent
of the vote and slightly more than half of the legislature. However, Moroz defected to
Yanukovych.
44. Andriy Konovalenko, “Баба Поразка Віктора Ющенка” [Baba defeat of Victor
Yushchenko], Ukrainska Pravda, January 26, 2010.
45. Serhii Leshchenko, “Порошенко і порожнеча” [Poroshenko and the vacuum],
Ukrainska Pravda, May 16, 2014, www.pravda.com.ua/articles/2014/05/16/7025568.
46. See the OSCE election reports available at https://www.osce.org/odihr/elections/
Ukraine (accessed July 2021).
47. In early February 2021, the government shut down three television stations owned
by an oligarch, Viktor Medvedchuk, with personal ties to Putin. This move was
supported by the US government as a “defense of [Ukraine’s] sovereignty and territo-
rial integrity” against “Russia’s malign influence” (Kalashnyk 2021).
48. Conflicts in Moldova in 1991–92 and Ukraine after 2014 were much more geograph-
ically confined than in Georgia, where civil war encompassed the whole country
in the early 1990s. In addition, Georgia, in contrast to Moldova and Ukraine, was
surrounded on all sides by nondemocratic states—Turkey, Russia, Azerbaijan, and
Armenia. Finally, Way (2015) argues that identity divisions between relatively equal
groups sustained political competition in Moldova and Ukraine, whereas the absence
of this kind of division in Georgia made the monopolization of power easier.
49. Between 1991 and 2021, Moldova witnessed four peaceful, institutionally regular
democratic turnovers (1996, 2001, 2016, and 2020) and Ukraine experienced three
(1994, 2010, and 2019), while Georgia experienced only one.
50. On these events, see Fairbanks 2004, 117; Karumidze and Wertsch 2005, 15, 39, 54;
and Mitchell 2004, 348.
51. The US government provided more than $1 billion in aid between 1992 and
2002, making Georgia one of the highest per capita recipients of aid in Eurasia
(Nichols 2003, 5).
52. See Dolidze 2007; “Viewing Georgia, without the Rose-Colored Glasses,” New York
Times, September 25, 2008. As Nodia (2005, 1) stated, “strengthening the state was
accompanied by certain setbacks in democratic freedoms.”
53. Joshua Kucera, “Georgia, Tajikistan, among Countries Most Dependent on U.S.
Military Aid,” Eurasianet, March 11, 2016, https://eurasianet.org/georgia-tajikistan-
among-countries-most-dependent-on-us-military-aid.
54. Margarita Antidze, “Georgian Tycoon Takes on Saakashvili,” Reuters, October 17,
2011, https://www.reuters.com/article/us-georgia-tycoon/georgian-tycoon-takes-
on-saakashvili-idUSTRE79G1CT20111017.
55. Lincoln Mitchell, “Could Georgia Be 2012’s October Surprise?,” RealClearWorld,
September 22, 2012, https://www.realclearworld.com/articles/2012/09/22/could_
georgia_be_2012s_october_surprise.amp.html.
56. See Liz Sherwood-Randall, “President Obama Meets with Georgian President Mikheil
Saakashvili,” The White House, President Barack Obama (blog), February 3, 2012,
https://obamawhitehouse.archives.gov/blog/2012/02/03/president-obama-meets-
georgian-president-mikheil-saakashvili; Vladimir Socor, “Georgia and the United
States: De-Alignment through Regime Change? (Part Two),” Eurasia Daily Monitor,
March 18, 2013, https://jamestown.org/program/georgia-and-the-united-states-de-

alignment-through-regime-change-part-two/; Lebanidze 2014, 211.
57. Socor, “Georgia and the United States: De-Alignment through Regime Change?
(Part Two)”; Giorgi Lomsadze,. “Saakashvili and Ivanishvili: Bonkers for Obama,”
Eurasianet, January 31, 2012, https://eurasianet.org/saakashvili-and-ivanishvili-
bonkers-for-obama; Cecire 2013, 247.
58. Josh Rogin, “Inside the Other Georgian Lobbying Effort in Washington,” Foreign
Policy, January 30, 2012, https://foreignpolicy.com/2012/01/30/inside-the-other-
georgian-lobbying-effort-in-washington/.
59. Cecire 2013, 237; “Saakashvili Comments on Ivanishvili,” Civil.ge, May 21, 2012,
https://civil.ge/archives/121944.
60. “Georgian Dream Shown the Exit Door,” Nezavisimaya gazeta, October 30, 2018, p. 5.
61. “The Fall and Rise of Georgia’s Most Watched TV Channel—Rustavi 2,” Jam News,
August 21, 2019, https://jam-news.net/the-fall-and-rise-of-georgias-most-watched-
tv-channel-rustavi-2/.
62. Ibid.
63. “Georgia Seeks Interpol ‘Red Notice’ for Saakashvili,” Civil.ge, August 31, 2014,
https://old.civil.ge/eng/article.php?id=27628.
64. Margarita Antidze, “Europe’s Human Rights Court Rules against Owners of Georgian
Pro-Opposition TV Channel in Ownership Row,” Reuters, July 18, 2019, https://www.
reuters.com/article/us-georgia-p olitics-tv-court/europ es-human-r ights-court-
rules-against-owners-of-georgian-pro-opposition-tv-channel-in-ownership-row-
idUSKCN1UD1YY.
65. Giorgi Lomsadze, “In Georgian Politics, It’s Millionaire vs Billionaire,” Eurasianet, August
7, 2019, https://eurasianet.org/in-georgian-politics-its-millionaire-vs-billionaire.
66. See Freedom House (2020b) and the detailed report on this by Transparency
International (David 2020). Furthermore, in early 2020, the government canceled
a major port project that Khazaredze’s company benefited from—a move that was
seen by some as punishment for Khazaredze’s political activity. See Giorgi Lomsadze,
“Georgia Cancels Contract for Black Sea Megaport,” Eurasianet, January 9, 2020,
https://eurasianet.org/georgia-cancels-contract-for-black-sea-megaport.
67. Georgian Dream won 90 of 150 seats.
68. “Lelo’s Khazaradze Speaks of Church, Queer Rights, Foreign Policy,” Civil.ge, June 8,
2020, https://civil.ge/archives/355161.
69. See, in particular, discussions of Kenya, Zambia, Mali, and Madagascar.
70. All successful opposition candidates from 1991 to 2016 ran as independents and
lacked any party organization backing them.
71. Kathryn Nwajiaku (1994, 435) argues that groups pushing the democratic transition
emerged out of the “splintering” of Kerekou’s patrimonial networks.
72. Benin’s armed forces were “the least coherent in Francophone Africa” and existed as a
“patchwork of conflicting political allegiances” (Decalo 1976, 53).
73. Soglo’s allies were not bound together by any political organization or “any grand
ideals beyond the immediate wrenching of power and sharing in the spoils of office”
(Amuwo 2003, 163).
74. The military attempted coups in 1992 and 1994—the latter of which was only halted
because of US intervention (Magnusson 2005, 223, 224).
75. In 1994, Kérékou created Action Front for Renewal and Development (FARD)
which won just ten seats in 1995 and relied on an extremely localized base of sup-
port (Bierschenk, Thioléron, and Bako-Arifari 2003, 163). As a result, he was forced
to cobble together a relatively loose and unstable coalition that suffered numerous
defections.
76. David Lewis, “Third Term Doubts Overshadow Benin Legislative Vote,” Reuters, April
22, 2015, https://jp.reuters.com/article/benin-politics-idAFL5N0XJ4TO20150422.
77. While Yayi ran as an independent, the Cowry Forces supported him in the legislature.
See “Benin Votes in Key Test for President Boni Yayi,” Agence France Presse, April 26,
2015; and Albert Trithart, “Mixed Results as Term Limits Put to the Vote on Africa’s
‘Super Sunday,’ ” IPI Global Observatory, March 28, 2016.
78. Tyson Roberts, “Here’s Why Benin’s Election Was a Step Forward for African Democratic
Consolidation. And Why It Wasn’t,” The Washington Post, March 22, 2016.
79. Freedom House, “Freedom in the World 2019: Benin,” https://freedomhouse.org/
country/benin/freedom-world/2019; Virgile Ahissou, “African Court Slams Benin’s
Treatment of ‘Chicken King’ Ajavon,” Bloomberg.com, March 30, 2019, https://www.
bloomberg.com/news/articles/2019-03-30/african-court-slams-benin-s-treatment-
of-chicken-king-ajavon.
80. Freedom House, “Freedom in the World 2019: Benin”; Ahissou, “African Court,” 2019.
81. “Benin MPs Adopt New Electoral Law,” BBC Monitoring Africa, September 4, 2018.
6
The Puzzle of Timor-Leste
Nancy Bermeo
When Timor-Leste voted for independence in 1999, the deck seemed stacked
against successful democracy.* The nation had a miniscule middle class, over
40 percent of the population lived on less than 55 cents per day (UNDP 2002),
fewer than 50 percent were literate (Millo and Barnett 2004), and life expec-
tancy was only forty-eight years (Karatnycky 2002). Only seventeen countries
on the planet had worse human development indicators (Harris and Goldsmith
2011). When it held its first national elections in 2002, Timor-Leste’s GDP per
capita was distressingly distant from the average for other nominally consoli-
dated democracies. Figure 6.1 illustrates this using a seventeen-year threshold
for consolidation (Svolik 2015, 717). Seen in the light of the vast literature linking
democracy’s longevity to economic development, Timor-Leste was the quintes-
sential outlier.
Social homogeneity is thought to advantage new democracies, but Timor-
Leste seemed disadvantaged on this dimension as well. Area specialists
pointed to at least thirty ethnic or tribal groups (Anderson 2001), sixteen
different languages, multiple dialects, and an East–West regional divide
(Palmer and McWilliams 2018). Observers saw it as “not one place but many”
(Shiosaki 2017, 55).
These socio-structural obstacles combined with historical factors that might
work against success. First, the continued presence of strong and legitimate tra-
ditional authorities at the village level could pose a challenging counterforce
to democratically elected leaders (McWilliams 2008; Swenson 2018). Second,
the nation lacked any previous experience with a working democracy. Quite
the contrary, when the fall of the Portuguese dictatorship made independence
possible in 1974, many local elites believed that “society was ill-prepared . . . to
engage in political activity” and that the few political organizations that existed
were merely “small cells” operating “in ignorance of one another” (CAVR 2005,
14, 23). The movements for national liberation that had emerged in Portugal’s
African colonies had never emerged. To make matters worse, the country’s tenta-
tive steps toward a democratic transition in 1975 had ended in a bloody internal
war between inchoate political parties. This armed struggle was quickly followed
by an invasion from Indonesia and then a brutal, twenty-four-year separatist war.
The Puzzle of Timor-Leste 161
5000 10000 15000 20000 25000

GDP Per Capita in 2011 US$
Timor-Leste
0
1950 1960 1970 1980 1990 2000 2010

Year
Timor-Leste Mean for Democracies Lasting 17+ Years
Figure 6.1. Transition year GDP vs. mean for existing democracies, 1945–2015
Sources: Bolt et al. (2018) and Marshall and Jaggers (2018)
Timor-Leste’s watershed wars produced a third raft of factors that might un-
dercut democratic success. Demobilized troops have often been harmful in new
democracies (Lyons 2004), and Timor-Leste’s democracy would have to be built
as thousands of “demobilized” troops re-entered a ruined economy with arms
in hand. Relatedly, the devastation wrought by decades of violence meant that
Timor-Leste would be creating a new democracy not simply in a new state but
in a weak and “mendicant state” (Shoesmith 2003, 234). Finally, FRETILIN, the
largest and most visible party associated with the new nation’s wars, had strong
historical ties with FRELIMO, the party that led what many deemed a hybrid
regime in Mozambique. This led observers to predict that if a lasting electoral
democracy emerged in Timor-Leste at all, it would very likely be a one-party-
dominant system.
Surprising Success
Despite the long litany of liabilities listed above, Timor-Leste has maintained
a democracy of surprising durability. The democratic regime that emerged
amid so much uncertainty in 2001–02, is still in place some twenty years later.
Profound challenges remain but, if Milan Svolik (2015) is correct in predicting
that democracies that survive this long have a breakdown risk of only one in two
hundred, Timor-Leste has already crossed a meaningful threshold.
Timor-Leste’s democracy is seriously challenged, but whether democracy re-

mains the “only game in town” or not, the longevity of its current regime is extraor-
dinary. Figures 6.2 and 6.3 illustrate the point, using a Polity Score of 6 or above as
the threshold for considering a country a democracy. Figure 6.2 shows that Timor-
Leste compares very favorably with former Portuguese colonies in Africa.
10
5
Polity IV Score
0
–5
–10
1975 1985 1995 2005 2015

Year
Timor-Leste Cape Verde Mozambique

Guinea-Bissau Angola São Tomé
Figure 6.2. Polity scores for former Portuguese colonies, 1975–2015

Source: Marshall, Gurr, and Jaggers 2016
10
5
Polity IV Score
0
–5
–10
1990 1995 2000 2005 2010 2015

Year
Timor-Leste Fiji Malaysia

Papua New Guinea Thailand Indonesia
Figure 6.3. Polity scores for post-1989 Asian democracies, 1990–2017

Source: Marshall and Jaggers 2018
Figure 6.3 shows that the durability of Timor-Leste’s democracy compares

very favorably to many of its Asian neighbors. Even in richer countries, such as
Fiji, Malaysia, Papua New Guinea, Sri Lanka, and Thailand, democratic spells
have been shorter lived.
Timor-Leste’s democracy confounds the common wisdom in terms of quality
as well as durability. Figure 6.4 shows how Timor-Leste compares to some of its
Asian neighbors on a widely used Civil Liberties Index. Figure 6.5 provides the
same case comparison with a widely used Liberal Democracy Index.
Closer case analysis justifies the positive assessments in these large data sets.
Knowledgeable observers report that the “country’s population is deeply engaged
politically” (Aspinall et al. 2018, 155). In fact, voter turnout for national polls
has averaged 78 percent since the founding of the nation with only a modest de-
cline over time. This figure matches that of South Korea and compares favorably
with those of much richer and more established democracies such as Germany
and Norway (IFES 2018). Most important, voters are participating in elections of
consistently good quality. Sérgio Vieira de Mello, of the United Nations, deemed
Timor-Leste’s first elections of a standard which could make “many democratic
countries . . . jealous” (King 2003, 751). The 2007 elections actually brought a
turnover in power, marking “the emergence of a genuinely competitive multi-
party system” (Leach and Kingsbury 2013, 22). The 2012 elections were deemed
“a resounding success” (Shiosaki 2017, 61–62), and the 2017 elections brought a
remarkable, second turnover in power. Timor-Leste is noted for a “lively public
1
.8
Civil Liberties Index
.6
.4
.2
0
1990 1995 2000 2005 2010 2015

Year
Timor-Leste India Indonesia

Thailand Malaysia Philippines
Figure 6.4. Civil liberties for Southeast Asia and India, 1990–2017
.6
.4
.2
0
1990 1995 2000 2005 2010 2015

Year
Timor-Leste India Indonesia

Thailand Malaysia Philippines
Figure 6.5. Liberal democracy for Southeast Asia and India, 1990–2017
sphere,” “vigorous competition among parties,” the absence of retail vote buying
(Aspinall et al. 2018, 153–54), and the fact that “parties do offer real choices” with
“some response to public interest” (Shoesmith 2013, 126). Borrowing Varshney’s
distinction we can conclude that Timor-Leste’s democracy appears both vibrant
and liberal (2015).
Solving the Puzzle
What explains this positive but perplexing outcome? For Timor-Leste, and
indeed for many other cases, the answer lies in the legacies of the armed con-
flict that preceded its democratic regime change. Ironically, armed conflict can
sometimes be helpful to democratic institution-building. Wars can disrupt the
foundations of authoritarian coalitions and lay the foundation for new coalitions
legitimated through popular electoral mandates. The foundation for a lasting
electoral democracy has traditionally had two basic elements: organized sets of
party elites who agree to compete for power through free elections, and a military
establishment that is consistently willing to support free elections and the au-
thority of elected leaders. Both of these institutional elements can and do emerge
without armed conflict, but they can also emerge as a result of internal war.1
Of course, internal wars do not always produce changes that are helpful to
democracy. The legacies of internal war vary with the nature and outcome of
the war itself. Which sorts of war are helpful and which are not? Quantitative
analysis of all new democracies emerging between 1946 and 2011 reveals that
revolutionary wars have a significant, positive effect on democratic durability;
that identity wars, in the aggregate, have no statistically significant effect on
durability in either direction; but that a small subset of identity wars appear to
produce democracies of notable resilience. Successful separatist wars which si-
multaneously generate both a new state and a new democracy produce highly
durable regimes. I call these breakaway democracies.2 They are few in number but
relatively resilient, and Timor-Leste is one of these.3
The Kaplan-Meir Plot in Figure 6.6 shows the survival probabilities of new
democracies with various conflict histories.
These patterns derive, in large part, from the divergent effects that different
sorts of conflict have on political parties and military institutions. Revolutionary
wars are fought for “the people,” a group without fixed boundaries. Identity wars
are fought for “a people” instead. Though identities are certainly malleable, the
relatively fixed social boundaries so often reinforced by identity wars can compli-
cate the creation of inclusive, and therefore viable, political party systems. They
can also complicate the creation of inclusive and non-interventionist armed
forces. Breakaway democracies have three advantages which make them better
able to cope with these complications. First, the act of warring with the “rump”
state can build new solidarities across divisions that were problematic before the
separatist victory. Second, the act of separation itself typically produces more
1.00
0.75
Survival Probability
0.50
0.25
0.00
0 20 40 60 80
Years
Unarmed Conflict Revolutionary
Identity Breakaway
Figure 6.6. Conflict history and democratic durability, Kaplan–Meier Plot

Sources: Boix-Miller-Rosato (BMR) dichotomous coding of democracy, 1800–2015; Cederman,
Wimmer, and Min 2010; Geddes, Wright, and Frantz 2014; Kalyvas and Balcells 2010; Marshall and
Jaggers 2018; Toft 2012; UCDP/PRIO Armed Conflict Dataset.
homogeneous polities. Finally, separation offers the opportunity to custom-craft

new institutions to lower the risks of democratic competition.4
This is simply a brief summary of a much broader project, but the argument’s
core and the quantitative evidence that sustains it suggest that the surprising
endurance and quality of Timor-Leste’s democracy derives, in large part, from
the institutional and ideational legacies of the wars that preceded its founding.
Historical analysis of the armed struggles that raged between the end of
Portuguese colonialism in 1975 and the beginning of Timor-Leste’s new democ-
racy in 2002 illustrates these points. A conflict history that included both a brief
war with revolutionary elements and a successful separatist war resulting in a
new state created the institutional and ideational landscape for a relatively robust
democracy.
Together Timor-Leste’s wars provoked four major changes that bolstered
successful democracy. First, they weakened the forces that undergirded dic-
tatorship by diminishing the ranks of the local actors who had formed the
anti-democratic coalition of the past. Second and relatedly, they strengthened
an inclusive national identity that not only replaced the diffuse dependency-
based identities of the past but also gave rise to a series of inclusive institutions
(and mindsets.) Third, the wars reinforced the infrastructure for a competitive
party system. They did this by producing a host of heroic leaders who were
willing and able to compete successfully in free elections, by creating the infra-
structure for several viable party organizations, and by attracting the external
resources to make the rewards of democratic competition greater than its risks.
Fourth and finally, Timor-Leste’s wars produced a professionalized, nonin-
terventionist, military establishment that supports the rule of elected leaders.
They did this in two ways: by bolstering the inclusive nationalist narratives that
eventually became the grounding for inclusive armed forces institutions and by
producing a range of leaders who could credibly link democratic institutions
with the coercive apparatus, giving the latter a credible stake in the new order.
The historical analysis that follows shows precisely how these consequential
changes took place.
From a Tentative Transition to Civil War
The revolution that brought democracy to Portugal in 1974 soon brought unwel-
come change to East Timor. Though a minister from the revolutionary govern-
ment in Lisbon quickly arrived to tell a hopeful crowd that “Timor will be what
the majority of people want it to be” (quoted in Dunn 2003, 64–65), the “people”
would not get to rule for nearly a quarter of a century. Two bloody wars and a
brutal dictatorship intervened.
The trouble to come was not immediately apparent. As soon as the Portuguese
eased restrictions on associational life, local elites established three proto-parties
representing a credible spectrum of ideologies. The Timorese Democratic Union
(UDT) was founded by Catholic professionals and assimilados, many of whom
were plantation owners and senior officials in the Portuguese colonial admin-
istration (Hicks 2016). The party was divided between a centrist, liberal wing
represented by Mário Viegas Carrascalão and an ultraconservative, fiercely anti-
communist wing led by Francisco Lopes da Cruz.
The Timorese Social Democratic Association (ASDT) situated itself on the
UDT’s left. Its founders were similar to the UDT’s founders in terms of social
background, and they too worked in government service, but they tended
to be less senior and, often, teachers. Like the UDT, the party embraced a
range of ideologies. A center-left group was led by Xavier do Amaral, a pro-
gressive Catholic who had studied theology in Macau. A more radical leftist
group was led by Mari Alkatiri, who had studied topography in Angola,
made contact with the MPLA, and become a Marxist (Magalhães et al. 2007).
Though the election of Xavier do Amaral as president suggested that mod-
erate forces in the party were probably numerically dominant, by September
1974 the more radical wing of the party changed the organization’s name to
the Revolutionary Front for an Independent East Timor (FRETILIN) and re-
vised its official program to focus much more explicitly on a social revolu-
tionary agenda.
The third party to be founded was the Popular Democratic Association of
Timor (APODETI). Its founding President, Arnaldo dos Reis Araújo, was a cattle
baron, but its organizers included landowners in other sectors, leaders of the siz-
able Arab community in Dili, and a number of liurai (local chiefs), who feared
that free party competition and an independent state would erode their consid-
erable power. Despite the social diversity of its founders, APODETI was more
ideologically homogeneous than its competition: the group was fiercely anti-
communist and single-minded about the goal of integration with Indonesia. In
fact, the party’s original name was the Association for the Integration of Timor in
Indonesia (AITI) (Hicks 2016).
APODETI’s focus on making Timor-Leste part of Indonesia was emblematic
of a major obstacle to a successful democratic transition. At both the elite and the
mass level, the people of Timor-Leste lacked a consistent and encompassing vi-
sion of national identity. They thus lacked the sense of national unity that Rustow
(1970) and others have found essential for viable democracy. Ambivalence
about nationhood was not confined to one party. A traditionalist, monarchist
party called KOTA sought fusion with Indonesia too, and another, smaller
party sought integration with Australia. Along with the small Labour Party,
the UDT initially sought federation with Portugal. Even FRETILIN, which had
sought independence from its inception, initially envisioned a gradual separa-

tion process. Benedict Anderson noted an “unsureness of identity” among “East
Timorese leaders” and concluded that in 1974–75 “true East Timor nationalism
was still quite thin on the ground” (2001, 237).
Unsettled national identity soon mixed with rabid anti-communism to
turn an incipient party system into a theater for civil war. Haunted by the pos-
sibility of independence under a FRETILIN government, Indonesian security
forces and APODETI leaders convinced right-wing anti-communists in the
UDT that FRETILIN would soon use Chinese arms and Vietnamese support
troops to turn Timor-Leste into a communist regime (Jardine 1999). After
meeting with a group of Indonesian generals, the ultraconservative presi-
dent of the UDT, Lopes da Cruz, joined forces with APODETI and launched
a coordinated, armed action throughout the country on August 11, 1975
(CAVR 2005). The pro-Indonesian forces lacked discipline and often executed
FRETILIN supporters on the spot (Dunn 2003), but FRETILIN’s ranks grew
quickly. It offered fierce resistance and swiftly drove the Indonesian-backed
coalition into West Timor. Though this first war lasted less than two months, it
killed between 1,500 and 3,000 people and displaced more than 20,000 others
(Hicks 2016; Taylor 1999).
The brief but brutal war left lessons in several quarters. Ordinary people
were “profoundly traumatized by the partisan conflict of 1975” and were left
“deeply apprehensive” about the reemergence of cross-party violence (Ingram
2018, 369). Party leaders in both the UDT and FRETILIN recognized the chaos
unleashed by armed men on both sides and openly admitted that they had “lost
control of the situation” (CAVR 2005, 31–40). Recognizing that the UDT had
politicized the upper ranks of the (still) colonial military and that FRETILIN
had politicized enlisted men, many East Timorese civilians and soldiers came
to recognize the danger of partisan division of the armed forces (CAVR 2005).
Finally, pivotal party leaders used the lessons of war to highlight the need for
inclusive nationalism: when he was inaugurated president of the FRETILIN
government in the war’s aftermath, Francisco Xavier do Amaral affirmed,
“After more than 400 years of . . . ignorance and massacres, what are we waiting
for? . . . we have to be the first and the last to resolve our problems. So, from this
day on, we all, yes, we all, will build our nation, Timor-Leste” (quoted in CAVR
2005, 56; emphasis added).
Tragically, the Indonesian government was already launching a military inva-
sion as these words were being spoken. Amaral’s presidency would last only ten
days while the occupation, and the war against it, would last over twenty-four
years. This second, very long, separatist war would reinforce the lessons emer-
ging from the civil war and leave additional legacies that eventually advantaged
the new nation’s new democracy.
The Democratizing Effects of the Separatist War
Being notoriously well armed, decidedly wealthier and much larger, Indonesia
expected a quick victory. It never came. The armed and unarmed supporters of
the resistance proved surprisingly resilient. FALINTIL, FRETILIN’s armed wing,
was nearly obliterated by the end of 1978, and defeat looked imminent as five
central committee members surrendered and FRETILIN president, Nicolau
Lobato, was shot dead. Yet the resistance fought on. In June 1980, FALINTIL
even managed an attack on the capital, Dili, “to show the world” that FRETILIN
still existed (CAVR 2005, 90). Though a notoriously brutal Indonesian officer
boasted in 1983 that his forces had reduced the resistance to “five hundred
members” with “one hundred weapons” (quoted in Kammen 2009, 73), the
Indonesians initiated a cease-fire in the same year. The costs of containing the re-
sistance through armed force were beginning to strain Indonesia’s dictatorship.
The resistance raised costs further by changing tactics. FALINTIL reorganized
into small, mobile guerrilla units. Meanwhile, resistance leaders began to invest
more in both clandestine work at home and diplomacy abroad (Robinson 2010).
Violence continued until 2001, but the course of the war eventually brought
changes that would be helpful to Timor-Leste’s new democracy.
Weakening the Anti-Democratic Coalition
The first change involved diminishing the power and ranks of the actors who had
poisoned party politics in 1974–75. Property owners helped derail democracy
in the mid-1970s, but the occupation benefited very few of them. Quite the con-
trary, landowning families were often targets of individual greed as local stooges
of the Indonesian military seized plantations (Kammen 2015). Coffee planters,
sandalwood producers, and other agriculturalists were forced to sell their
products at low prices when Indonesian military officers established commercial
monopolies in the vacuum created by their slaughter of ethnic Chinese (Dunn
2003; Jardine 1999). Ironically, the anti-democratic coalition built around the
fear of communist expropriation found many of its members subject to expro-
priation from ardent anti-communists. As Figure 6.7 illustrates, property rights
were abysmally weak until after democracy was established.
The Special Forces Command sent to “preserve order” in Timor-Leste treated
the province as “their special property” (Dunn 2003, 299–300), making the prov-
ince “a virtual fiefdom” (Aspinall and Berger 2001, 1011). As a consequence,
an increasing number of aristocratic and mestizo planter families came to play
“a leading role in the resistance” (Kammen 2015, 135). The harm done to local
property owners explains why even Indonesia’s local development programs
5
Property Rights Score
4
3
2
1
1975 1985 1995 2005 2015

Year
Timor-Leste Indonesia
Figure 6.7. Property rights in Timor-Leste vs. Indonesia, 1975–2015

backfired and eventually “encouraged the emergence of a ‘nationalist conscious-

ness’ ” (Mendes 2005, 205; my translation).
The nature of wartime violence weakened the foundations of the authoritarian
party coalition as well. The Indonesian Army used indiscriminate violence—
often killing APODETI and UDT supporters in their homes and in the streets—
especially in the early weeks of the invasion. This and the military’s engagement
in extensive robbery and rape produced a sharp decline in popular support for in-
tegration as early as 1976 (Dunn 2003). After the slaughter of university students
in the Santa Cruz Cemetery in 1991, insiders report that “Apodeti people started
to side with the resistance” en masse and that the organization began “to drown”
as a party (Anderson, Djati, and Kammen 2003, 21).
The UDT was shaken too as members scrambled to distance themselves
from the dictatorship. A number of party members were so revolted by the af-
termath of invasion and annexation that they attempted a coup in April 1976.
João Carrascalão, who had led the armed Anti-Communist Movement against
FRETILIN in 1975, soon became a prominent organizer for the resistance in
Australia, even aligning with FRETILIN activists (Durand 2016). Another UDT
founder, Mário Carrascalão, served as the Indonesian-appointed governor of
Timor-Leste but met with FRETILIN leaders several times, framed himself as
part of the resistance in 1983, and openly joined the broad resistance front in
1988 (Fernandes 2011).
The Catholic Church was never part of the dictatorial coalition but any
officials who might have sympathized were soon silenced by the overwhelming
support the local church gave to the resistance and to FRETILIN’s sympathizers
specifically. The local church was no longer “a bastion of the Portuguese colonial
system” (CAVR 2005, 93). From the war’s beginning it served as a “sanctuary” for
ordinary people, a “vital link to the outside world” (ibid.) and a bridge between
legality and clandestinity for the opposition (ibid., 99–104).
The enlisted men who were conscripted into what became the Indonesian-led
army constituted a final coalition group that changed during the war. As the con-
flict wore on, hundreds of Timorese deserted Indonesian battalions with their
weapons in hand and joined the resistance instead. As the occupation army’s
brutality grew, thousands went into hiding to avoid conscription altogether
(Dunn 2003).
Strengthening National Identity
Changes in group behavior were intimately tied to the fact that war strengthened
a sense of national identity. Years of lonely struggle against Indonesian annexa-
tion gave the people of Timor-Leste a strong sense of who they were and who they
were not. The “different visions of East Timorese nationalism” that were problem-
atic in 1975 were largely “subsumed” by the resistance struggle itself and gave rise
to a national identity that was independent but also inclusive (Aspinall and Berger
2001, 1012). Annexation and the war against it made the “colonized” more aware
of their commonalities and their ability to “emancipate themselves” by working in
concert (Anderson et al. 2003). The “ ‘occupied-occupier’ binary . . . [still] frames
national identity today” (Arthur 2019, 10), trumping alternative identities based
on ethnicity, religion, language and, to lesser extent, region. As FRETILIN leader
Mari Alkatiri himself put it, “national cohesion was built upon opposition to the
occupant” (quoted in Shoesmith 2003, 242–43).
Invasion opponents within the local clergy provided key mechanisms for
brokering cohesion and for bolstering inclusivity. They were uniquely suited
to do so because (regardless of party) “almost all” of the political leaders who
spanned the period from 1974 through the occupation had studied in the same
Jesuit schools (Cristalis 2009, 23), and even Alkatiri, a Muslim, had published
political criticism in the local Jesuit magazine. Local clerics drew on these long-
standing networks to initiate dialogue and cooperation across party boundaries.
In 1982, for example, the highest Catholic official in the province met secretly
with FALINTIL’s leader, Xanana Gusmão, to highlight both “the need for na-
tional unity” and “his understanding of the Resistance” as a purely “nationalist”
rather than a leftist struggle (CAVR 2005, 97–98). The following year, a priest
who had invited Gusmão to speak in his church accompanied him to a meeting
with Mário Carrascalão, the Indonesian-appointed governor.
The role of the local church as a mechanism for creating a national iden-
tity extended beyond elites. Benedict Anderson argues that “a popular
Catholicism . . . emerged as an expression of a common suffering, just as it did in
nineteenth-century Ireland” and that the emphasis on “commonality” became a
substitute “for the kind of nationalism . . . which comes from print-capitalism”
(Anderson 2001, 238).
The fact that the local church chose Tetum as the language of the liturgy (de-
liberately eschewing both Indonesian and Portuguese) had “profoundly nation-
alizing effects” (Anderson 2001, 238). These connections between war, language,
and identity construction were obvious to FRETILIN leader José Ramos-Horta
years before he and Bishop Carlos Belo would win a Nobel Prize for their role in
the resistance:
Independence is now the desire of everyone, including those who had

illusions . . . years ago. . . . Indonesia’s efforts to “pacify” us have failed, as have the
attempts to destroy Timorese cultural identity. Throughout the country . . . there
is a tremendous cultural movement. . . . The Tetum language, which was once
spoken by slightly over 50% of the population, is now used by everyone and has
gained tremendous strength and vitality. (Ramos-Horta 1987, 205)
Despite the role that Catholic clerics played in the strengthening of Timor-
Leste’s national identity, the identity itself was not religiously exclusive. To the
contrary, the war of liberation gave rise to what Tudor and Slater call “inclusive
nationalism,” a form of nationalism that explicitly challenges “colonial practices
of divide and rule” and instead serves to mobilize “active, direct support across
class . . . ethnic and religious divisions” (Tudor and Slater 2016, 28–29). This is
one of the second war’s most important legacies. As Tudor and Slater have shown
convincingly, inclusive nationalism has been a powerful “source of democratic
strength” in countries such as India (and later, Indonesia) where democracy de-
veloped against the odds (Tudor and Slater 2016, 28; Tudor 2013).
National identity in Timor-Leste draws on the highly inclusive concept
of Mauberism, an identity frame that began to develop in the last years of
Portuguese colonialism. Mauberism glorifies an indigenous “brotherhood”
united in suffering from external oppression but elevated by ancient traditions
of “self-rule” (Myrttinen 2013, 214), fearlessness, “cooperation[,]‌and consensus”
(Cristalis 2009, 24). The celebration of Maubere identity enabled FRETILIN, and
later, the resistance as a whole, to “Timorize the idea of ‘the people’ (in a nation-
alist sense)” (Shoesmith 2003, 239).
In keeping with the prototype of “inclusive nationalism,” diversity is intrinsic
to the definition of who “the people” of Timor-Leste are. In 1986 Gusmão himself
wrote that independence was “the only path” for the “preservation of the identities
that distinguish our people” (Niner 2000, 124). Specialists observe “a national
identity that preserves and promotes cultural diversity” (Sousa 2001,194); “a
strong sense of aggregated nativeness that infuses [various]Timorese identities
with a sense of unity” (Palmer and McWilliams 2018); and a “general and diffuse
nationalism” in which “diversity” is not only a key component “of identity” but of
“national dignity” as well (Sousa 2001, 193–94).
The new inclusive nationalism emerging from war explains why—despite
deadly intimidation—nearly 80 percent of the people of Timor-Leste chose in-
dependence over autonomy when they were finally allowed to vote on the issue
in 1999. Timor-Leste’s second war had eroded and subsumed the divisions that
facilitated its first war, and had given rise to an inclusive nationalism instead.
Bolstering a Competitive Party System
In addition to bolstering inclusive nationalism, separatist war produced the

leaders and organizations essential to a competitive party system. By the time of
independence, Timor-Leste had a broad array of leaders who were willing and
able to win free elections. Some emerged on the battlefield, others in the exile
community. Still others emerged in the arenas the occupiers established as means
of control: including the sham local government institutions and the expanded
student community. Each leader became visible as a result of the challenges and
tactical conflicts inherent in waging a decades-long resistance war—and each
became associated with a party organization as Timor-Leste’s new democracy
developed. Thus, by the time of its founding elections in 2002, the cores of sev-
eral competitive parties had developed in Timor-Leste, producing a party system
that was balanced enough to avoid single-party domination.
FRETILIN was Timor-Leste’s most visible party as its founding elections
approached, and the nature of the war explains why. FRETILIN managed to con-
fine the invasion force to only 30 percent of the province for over two years, and
used liberated zones to expand its base, not just by keeping people alive but by
promoting literacy, agricultural cooperatives, healthcare, and participatory de-
mocracy (Capizzi et al. 1976; CAVR 2005; Taylor 1999). Its “great strength . . . was
to never use violence against civilians” (Durand 2016, 119).
The resistance was harboring some 450,000 civilians when Indonesia closed in
on the liberated zones (Durand 2016). But the experience of governance had al-
ready ensured FRETILIN “support in most regions” (Taylor 1999, 96) and forced
party cadres to bureaucratize and gain administrative skills.
Later watershed moments of the war worked to FRETILIN’s advantage as well.
When the Indonesians interned over 300,000 civilians to cut resistance support
(CAVR 2005), FRETILIN exploited the camps to recruit even more (Budiarjo
and Liong 1984). When the occupiers called a cease-fire after recruiting locals
into the Indonesian Army, FRETILIN successfully lobbied recruits to defect in
droves (Kammen 2009). When the death toll for men skyrocketed, FRETILIN
recruited women (Loney 2018; Niner 2015).
These wartime successes help explain why FRETILIN leaders were willing to
compete in fair elections. As a militant told an interviewer in 1992, “politically,
we [have] won” already (Jardine 1999, 75). Positive prognostications were appro-
priate. FRETILIN won the new nation’s founding elections with 59 percent of the
vote, but it did not become a hegemonic party because war had helped produce
powerful electoral alternatives.
A first set of alternatives emerged from within FRETILIN itself. This is because
a long and intense struggle with a better-armed power inevitably raises tactical
dilemmas. Different leaders responded differently to a range of dilemmas and in
so doing laid the foundation for different organizations.
The first fissure with organizational consequences involved Xavier do Amaral,
FRETILIN’s founding president. Recognizing that Indonesia’s 1977–79 encir-
clement campaign would inevitably kill thousands of innocent civilians under
FRETILIN “protection,” Amaral led a group back to his home region and began
to negotiate a local truce. Other FRETILIN leaders declared this treasonous and
had him arrested (Taylor 1999). When the war ended, he organized the ASDT,
an alternative party attracting strong support in his home region, and among
“older . . . revolutionaries who had grown disillusioned with FRETILIN” (King
2003, 754).
A second, more powerful party grouping emerged from tactical changes ini-
tiated by Xanana Gusmão, the charismatic head of FALINTIL. His eyewitness
experience of the war gave him a vision that many exiled FRETILIN leaders did
not share. It was a more pluralist, more inclusive vision of both the resistance and
the regime that should emerge after victory.
Beginning in 1983, he insisted that FRETILIN “must mobilize people who
[seek] an Independent Country,” regardless of their ideology, and that the UDT
and FRETILIN should consider “a course of National Unity” to secure indepen-
dence and peace (quoted in Kammen 2009, 87).
In December 1987, he rejected Marxist revolution altogether and made his
commitment to democracy explicit in a seminal speech that was taped and
shared worldwide (Weldemichael 2013). He stated clearly, “We have come to
recognize through these long years that what inspires the people of East Timor
to struggle is not the making of the revolution . . . [but the desire] to liberate
the Homeland . . .and . . . to live freely and independently. . . . I publicly declare
my total and wholehearted rejection of those doctrines that promote suppres-
sion of democratic freedoms in East Timor” (Niner 2000, 134–35). In a fur-
ther effort to signal his inclusive approach, Gusmão resigned from the party
and declared FALINTIL “a national army” officially separate from FRETILIN

(Niner 2000, 131–35).
Gusmão’s turnabout met with fierce condemnation from party leaders, but
his charismatic authority could not be challenged. His success in founding an
inclusive resistance coalition in 1988 meant that FRETILIN was no longer “the
party of the resistance” (Weldemichael 2013, 203) and that the war would leave
multiple heroes and multiple organizations in its wake. By 2007, a party founded
by Gusmão, the CNRT, would win control of government through the nation’s
second parliamentary elections.5
The war of resistance had effects on actors outside of FRETILIN as well,
and thus left still other party organizations in its wake. These represented a
broad range of ideologies. The Social Democrat Party (PSD) was founded by
Mário Carrascalão, who drew on his visibility as provincial governor and then
resistance-front activist to create what became FRETILIN’s major opponent
on the right and represent the type of reformed conservative party that Dan
Ziblatt has convincingly associated with stable democratic trajectories in Europe
(Ziblatt 2017). A resurrected and reformed UDT, led by João Carrascalão, served
the same purpose in representing a moderate but unambiguously democratic
center-right (Shoesmith 2013). Both parties emerged from a strengthened sense
of national identity and the deep disillusionment brought on by war with an au-
thoritarian occupier (Anderson et al. 2003).
Another competitive party that emerged as a direct result of Timor-
Leste’s conflict history is the center-left Democratic Party (PD). The PD was
founded by Fernando de Araújo, the heroic and (once imprisoned) leader of
RENETIL—a student resistance movement that began among East Timorese
students in Indonesia. Although the PD came into being less than three months
before the founding elections, it polled remarkably well, garnering 8.7 per-
cent of the vote nationwide. It drew (and draws) its support from the younger
cohort of the resistance who worked in common cause with FALINTIL “to
make the territory ungovernable” (Leach 2016). Many had shown themselves
“ready to risk the wrath of a ruthless military” in demonstrations and clandes-
tine activities of great consequence (Dunn 2003, 297). All were alienated by
FRETILIN’s exiles whom they “perceived as arrogant, privileged, and authori-
tarian” and likely to claim a disproportionate share of power and employment
(King 2003, 755). Suspicious of the Marxist exile group that had come to be
known as “the Maputo clique” (Croissant 2008, 656), the PD, like other new
parties, was born of the strengthened national identity and commitment to
democracy that emerged as a legacy of war with an authoritarian occupier
(Myrttinen 2013). The breadth and strength of the parties emerging in the
wake of the separatist war led to a party system that was not simply viable but
diverse and competitive.
Producing a Noninterventionist Military
The nature of Timor-Leste’s wars had profound effects on its armed forces as well
as its political parties, producing a professionalized, noninterventionist military
that has also been essential to democratic durability. The positive outcome was
not inevitable. The civil war, the Indonesian invasion, and the war that followed
left a regional division within the coercive apparatus that caused a serious secu-
rity crisis in 2006 and might have reversed democratization altogether. However,
the inclusive nationalist ideology strengthened by the war plus a series of inclu-
sive institutions associated with the resistance helped resolve the crisis and even
dampen regional divisions (Bexley and Tchailoro 2013; Braithwaite et al. 2012).
The professional and noninterventionist nature of Timor-Leste’s military has
long been recognized. The military was seen as “unthreatening” when democ-
racy first began (Smith 2004, 157) and is still generally viewed as “an instrument
rather than an actor in politics” (Mietzner 2012, 13; see also Croissant and Kuehn
2009). The fact that a military coup has never even been attempted in Timor-
Leste bears this out and sets the case apart from several of its Asian neighbors.
There can be little doubt that the professionalization and the political neu-
trality of Timor-Leste’s military emerged as a result of conflict history and that
both qualities were widely recognized before the nation’s founding elections.
FALINTIL’s prestige as a professional fighting force began with its role as the
triumphant underdog in the 1975 civil war (CAVR 2005), but its warriors were
elevated to truly heroic status during the war of resistance as participants in a
“David-and-Goliath battle” (Cristalis 2009, 4).
The elevated image of the armed forces was easily joined to the image of being
“above politics” among locals, but the professionalism of the armed forces was
obvious to outsiders as well (Robinson 2010, 146–47). Military officers from
abroad assessed FALINTIL as decidedly “more professional” than its (better
funded and US-trained) Indonesian counterpart even before the democratic
transition formally began (ibid.). FALINTIL’s “extraordinary discipline” (Dunn
2003, 297) and exemplary behavior in the face of horrid provocations from anti-
independence militias in the aftermath of the referendum confirmed its profes-
sional image further (Ballard 2008; Robinson 2010).
The armed forces’ nonpartisan stance, like the array of political parties
described above, was very much a product of changes occurring during the
course of the resistance war. Though FALINTIL was founded as the armed wing
of FRETILIN, it eventually operated with extraordinary “organizational au-
tonomy” from the party (Sindre 2016, 504). Autonomy was a necessity at first.
Many FRETILIN leaders were killed as the war began, and the distance of the
exiled party leadership from the war itself diminished their influence consider-
ably (Weldemichael 2013), leading, as one exile recalls, to nearly “total silence”
until after 1982 (Gama 1995, 102). But FALINTIL’s autonomy from FRETILIN
was due to ideas too. Gusmão’s decision to broaden the resistance coalition made
it essential that FALINTIL was “rebadged as a non-partisan nationalist army”
(Ingram 2018, 368), and the move worked. FALINTIL was effectively and per-
manently separated from FRETILIN control (Shoesmith 2013).
There was certainly a strategic rationale for the separation: since FALINTIL
was fighting a guerrilla war without “any significant external funding” (Morjé
Howard 2008, 263) the broadening of its domestic resource base was essential
to its survival, but Gusmão laid out detailed ideological reasons for the sepa-
ration as well. In a public letter he insisted that FALINTIL would be “outside
the party-political game” altogether and dedicated instead “to the defense of the
homeland of everyone” and to the defense of a constitution that guarantees “indi-
vidual and collective freedoms and the respect for the interests of all citizens and
social classes in East Timor” (Niner 2000, 132–33; emphasis added).
As the neutral defender of the freedoms of “all citizens,” FALINTIL became
an institutional agent of the inclusive nationalism that developed during the
war. Unfortunately, its success in this role would be severely tested in 2006, be-
cause of wartime legacies that worked against democratization. Prime among
these were: the masses of veterans seeking employment in a ruined economy;
a police force drawn primarily from the Western region where FRETILIN was
least popular; the veterans of an armed group led by Cornelio Gama which had
broken away from FALINTIL in 1985 and had aligned with FRETILIN’s Interior
Minister, Rogério Lobato (Shoesmith 2003); and, finally, the longstanding per-
sonal divisions between the new nation’s two elected executives, FRETILIN’s
Prime Minister Mari Alkatiri and President Xanana Gusmão.
These potential fault lines opened wide in the spring of 2006. Though
FALINTIL had been transformed into the national armed forces and renamed
the F-FDTL in 2001, it had been cut dramatically in size. Veterans with seniority
were allocated most of the positions in the first battalion formed and, because of
the war’s geography, they came disproportionally from the east of the country
(Boyle 2014). The police force (PNTL) was also restructured, but there mostly
westerners were rehired. Eventually, wartime regional and political differences
became explosive.
The security crisis erupted after westerners within the F-FDTL signed a
public petition charging the military with discrimination and then abandoned
their posts after deeming the government response inadequate (United Nations
2006). Though protests were peaceful at first, after several hundred men were
dismissed, deadly riots led by Gama, a police major named Alfredo Reinado,
and others mobilizing regional animosities spread throughout the country. As
the death toll rose, the government invited troops from Australia, New Zealand,
Malaysia, and Portugal to help restore order. The prime minister was forced to
resign but was replaced lawfully by Nobel laureate José Ramos-Horta. The nation
held free and fair elections for a new government as scheduled and produced a
peaceful transfer of power.
Timor-Leste’s democracy survived its security crisis through the skillful use
of inclusivity. Four mechanisms stand out as important. The first is the semi-
presidential system adopted from Portugal (which served as a key geographic
base for the exiled resistance and a center for institutional diffusion). Semi-
presidentialism allowed one major leader to check the inadequacy of another
while giving two sets of citizens representation in the new nation’s executive
(Feijó 2016).
Gusmão used an inclusive executive institution to democracy’s advantage, but
the skillful use of inclusivity extended to the party system too. Several of the po-
tentially “anti-system” movements that were problematic during the first elected
government were successfully integrated into the party system during the second.
Some movement leaders joined the Social Democratic Party, others joined the
Democratic Party, and still others formed new parties such as UNDERTIM
(which, despite its leaders’ previous hostility to Gusmão, was invited to join the
ruling coalition). Timor-Leste’s party system has thus far managed to absorb po-
tentially disruptive movement activity and thus meet one of the major challenges
for new democracies (Bermeo and Yashar 2016b).
A third inclusive means of coping with the divisions provoking the crisis in-
volved the creation of institutions in which the police and the military work to-
gether. In a peacetime version of the coalition formation developed during the
resistance war, a joint military-police command has been established in the
hope that “collaboration” will enable the “reconciliation of previous enmities”
(Wilson 2012, 193). Though the fusion of functions under control of the military
contradicted the advice of foreign actors, it has become a point of “nationalistic
pride” (ibid.). It accompanies a whole series of public gestures in which gov-
erning elites use the wartime experience as a means of strengthening common
bonds—such as having conflict protagonists sign a highly publicized “peace
agreement” and involving previously warring groups in “peace demonstrations”
to commemorate the Santa Cruz massacre (Scambary 2011).
Finally, the national budget was used inclusively as well. Gusmão might have
used solely punitive measures against those who mobilized against “Xanana’s
Boys” in 2006, but he organized a series of highly inclusive grants and pension
programs instead—and these too have been widely credited for dampening dis-
order (Kingsbury 2014; Roll 2014).
Ironically, the emergence and resolution of the crisis may enable the
armed forces to be even more supportive of democracy in the future. Security
scholars note that “the crisis led to an improvement in East Timor’s security
outlook” (Tansey 2009, 107), that “institutional rivalries between the police
and the army . . . [have] receded,” that “regionalist . . . divisions” have “virtu-
ally disappeared,” and that coups are “unlikely” because civilian leadership
enjoys “charismatic authority” throughout the senior ranks of the defense forces
(MacQueen 2015a, 763; see also Harris and O’Neil 2011).
It is noteworthy that even at the height of the crisis, no viable coup coalition
emerged within the military. This provides a convincing illustration of the lasting
“acceptance for civilian supremacy” among the nation’s armed forces (Kammen
2011, 127). Specialists argue that this acceptance stems from the early Marxist-
Leninist orientation of the resistance movement (Scambary 2011). But the ac-
ceptance of elected civilian authority is likely to have come from a later period
when the armed resistance embraced democracy instead. Over time, the accept-
ance of elected civilian authority was reinforced by other factors too. First, the
example of military rule furnished by the Indonesians provided no grounds for
emulation—quite the opposite.6 Second, the heroic status that the armed forces
enjoyed among citizens and politicians alike made it unlikely that democratic
institutions would ignore the organization’s needs. Finally, the fact that military
heroes have often been elected to high office means that the armed forces cur-
rently have easy access to power and thus, no need to seize it unlawfully.
Alternative Hypotheses
Clearly, the legacies of Timor-Leste’s wars have contributed to democratic dura-

bility in important ways. But before concluding, it is useful to weigh two alterna-
tive explanations for success.
A first involves oil and gas revenue from the seabed between the island of
Timor and Australia.7 Timor-Leste has indeed benefited from natural resource
income in a variety of ways. Its use in the funding of the veterans’ pension system
is a case in point (Roll 2018).8 Yet it would be wrong to accord this revenue de-
cisive weight in determining democratic success. After all, a great deal of careful
scholarship finds that the connection between oil revenue and lasting democracy
is strongly negative (Ahmadov 2014; Aslaksen 2010; Ross 2001). In the specific
case of Timor-Leste, oil revenue has sometimes outpaced state spending capacity
(Lundahl and Sjöholm 2007), always fallen far short of expectations for job cre-
ation, sometimes focused on the needs of imagined tourists rather than those
of the overwhelmingly rural population (Nygaard-Christensen 2016), and often
fostered communal conflicts (Bovensiepen and Nygaard-Christensen 2018).
Thanks to natural resource revenue, Timor-Leste is now, technically, in the
lower band of middle-income countries, but this categorization belies the fact
that half the people in Timor-Leste still live below the two-dollar-a-day pov-
erty line and that the nation remains among the poorest countries in the world
by several measures (Kingsbury 2018). Thus, country experts are right to view
Timor-Leste’s resource revenue as “both destructive and . . . productive” for po-
litical institutions (Bovensiepen and Nygaard-Christensen 2018, 414). It may
have sometimes contributed to democracy’s durability but was certainly not its
sufficient or its principal cause.9
A second and seemingly more powerful explanatory hypothesis involves
assistance from abroad. International assistance was not just massive but
long lasting: the UN spent over $52 million on the 1999 independence ref-
erendum and twenty nations deployed 11,000 troops to combat the terror
unleashed by Indonesian-led militias in its aftermath. After order was
restored, the UN established the United Nations Transitional Administration
in East Timor (UNTAET), a costly authority with “the broadest mandate of
any peacekeeping mission before or since” (Ingram 2018, 366). It involved
9,150 military personnel, 1,640 police, 1,670 international civilian staff, and
nearly 2,000 local staff—all in an area the size of Connecticut (Morjé Howard
2008, 260). Support from abroad continued years after democratic elections
began. In fact, the UN did not secure its “metaphorical ‘exit visa’ ” until
2012, after Timor-Leste held its third successful set of legislative elections
(MacQueen 2015a, 762).
It is tempting to conclude that the success of Timor-Leste’s democratization
is due to foreign assistance rather than to the legacies of war, but a closer look
at the evidence suggests that this is incorrect. First, we have to recognize that
most of the foreign assistance received by Timor-Leste was itself a legacy of
war. The massive aid for reconstruction was allocated because war had turned
the territory into what a British diplomat aptly described as “hell on earth”
(Durand 2016, 173). The massive peacekeeping missions were deployed pre-
cisely because the peace had not been kept. In this respect, foreign assistance
is not an alternative to my argument but an intrinsic element of it. To the ex-
tent that aid had a positive effect, it illustrates the positive consequences of
armed conflict.
Yet there are other reasons to reject foreign assistance as the primary expla-
nation for Timor-Leste’s success. To start with, quantitative analysis suggests
that there is no clear association between foreign assistance and the durability
of post-conflict democracies in general. Table 6.1 presents hazard models based
on data for all new, post-conflict democracies emerging between 1946 and 2011.
Controlling for the major structural, institutional, and regional variables associ-
ated with democratic longevity, plus conflict type, we find that the likelihood of
breakdown diminishes with GDP, with GDP growth, and (as predicted earlier in
this chapter) with a revolutionary conflict history. But we also find that neither
aid per capita (models 1–3) nor UN Peacekeeping (models 4–6) has a significant
effect in either direction.
Table 6.1. Aid, Peacekeeping, and Democratic Durability
Model 1 Model 2 Model 3 Model 4 Model 5 Model 6
Aid per 0.145 0.500 0.075

capita (SL) (0.292) (0.326) (0.399)
UN peacekeeping –0.989 –0.518 –0.448
(0.618) (0.695) (0.531)
Revolutionary –1.655* –1.886** –1.533* –1.894**
conflict (0.694) (0.665) (0.657) (0.638)
Identity conflict 0.339 0.389 0.214 0.435
(0.395) (0.367) (0.402) (0.377)
GDP per capita –1.973*** –1.829*** –1.863*** –1.750**
(SL) (0.353) (0.550) (0.368) (0.551)
GDP growth –0.308** –0.341** –0.274* –0.322*
(SL) (0.112) (0.115) (0.107) (0.113)
Trade openness 1.094* 1.097* 0.963* 1.016
(SL) (0.449) (0.547) (0.459) (0.542)
Ethnic 0.482** 0.459**
fractionalization (0.162) (0.152)
PR –0.546 –0.411
(0.356) (0.414)
Presidentialism –0.525 –0.539
(0.315) (0.301)
Region fixed No No Yes No No Yes
effects
Country clusters 97 94 94 99 95 95
Number of 2,363 2,266 2,266 2,394 2,279 2,279
observations
Note: A country-year was coded as democratic if it scored 6 or above on the Polity IV index. Standard
errors were clustered on country. Coefficients estimated from Cox proportional hazards models
using the Efron model for ties. Proportional hazard tests at the 0.05 threshold were used to check for
Time Varying Covariates. SL indicates a variable is standardized (S) and lagged (L) by 1 year.
* p < .1; ** p < .05; *** p < .01
In Timor-Leste specifically, the effects of foreign resources on successful de-

mocratization were decidedly mixed. Foreign troops were critical to the resto-
ration of order after the independence referendum and again in 2006–7. This
is why Timor-Leste is so frequently cited as an example of successful interna-
tional intervention (Morjé Howard 2007). But the role of UNTAET and other
international assistance bodies in promoting democracy in Timor-Leste is hotly
contested. They were deemed by some to be the cause of dangerous intra-elite

conflict (King 2003) and were sometimes viewed as “profoundly undemocratic”
(Caplan 2005, 171) or even “absolutist” themselves (Chopra 2002, 979).
Criticism was not ubiquitous, but the failure of international assistance was
sometimes more obvious than its success. The United Nations Mission to East
Timor’s (UNAMET’s) attempt to run the independence referendum in an or-
derly and secure environment exemplifies the point. Though UNAMET agents
were determined to quell regional violence by establishing a watchful presence
throughout the territory, running extensive voter education programs, and
arranging that only aggregate rather than district referendum results would
be reported (Shiosaki 2017), anti- independence paramilitaries killed over
2,000 people anyway. UNAMET’s staff and offices became targets themselves,
whole villages were razed to the ground, thousands had to flee their homes, and
70 percent of public buildings were destroyed (CAVR 2005). The international
community’s failure to provide sufficient security, despite “unassailable” evi-
dence that it was desperately needed, left Timor-Leste in even worse condition
after the balloting than before (Martin and Mayer-Rieckh 2005).
After independence, the UN’s failed attempts at police reform contributed de-
cisively to the 2006–7 security crisis. Though the development of “a credible, pro-
fessional and impartial police service” was a centerpiece of the UN intervention,
its officials made several highly consequential errors along the way (Hood 2006),
including relying excessively on the advice of Indonesian police officials, hiring
hundreds of inevitably distrusted Indonesian police veterans for the “new” na-
tional police force (Arnold 2009), excluding “key stakeholders” (Powles 2015,
212), and failing to consult with the local population in general (Chopra 2002;
Hood 2006).
In addition to suffering from “all but non-existent” civilian oversight (Hood
2006), the PNTL had deeply conflictual relations with the national armed forces.
These conflicts reflected the widespread perception that the PNTL was con-
trolled by and biased toward people from the western districts and a related
perception that the donor community was overfunding the police and under-
funding the military (Hood 2006). Though an expert outside evaluation team
concluded in 2003 that the UN police program had produced an institution
that was both “weak” and “unsustainable” (International Policy Institute 2003,
102), peacekeeping forces were withdrawn in May 2005 anyway. The with-
drawal proved “catastrophically premature” (MacQueen 2015b, 692), leading to
“crippling chaos” and the greatest political crisis in the new democracy’s history
thus far (MacQueen 2015a, 765).
Beyond evidence that assistance policies sometimes threatened democratic
longevity, there are other reasons to question the relative strength of assistance
effects. One of the most powerful relates to sequencing. The extraordinary
external resources described above were forthcoming in large part because

Timor-Leste’s political elites had made a credible commitment to democracy be-
fore the transition even got underway. This is one of the reasons why a senior
UNTAET official concluded that the new nation offered “conditions for success
rarely available” to peace missions elsewhere (Chopra 2000, 28). As Rebecca
Strating puts it, democratic values were not “simply imposed on East Timor.”
A new, “state-based identity . . . formulated around principles of pluralism [and]
democracy . . . developed during the independence movement” and then “jus-
tified and sustained it” (Strating 2016, 3; emphasis mine). Kammen argues,
similarly, that the amalgam of new organizations, political discourse, and pop-
ular activism that emerged as part of the resistance to the Indonesian occupa-
tion constituted “a perverse form of democratization” in itself (Kammen 2015,
159). After independence, foreign advisors “guided institutional design,” but
institutions “evolved in response to local political drivers” (Ingram 2018, 365).
International efforts to build and protect democracy in Timor-Leste did con-
tribute, in part, to the durability of Timor-Leste’s democratic institutions but as
a secondary influence. As both Thomas Carothers (1999) and Richard Caplan
(2005) have argued, even the most successful democracy-promotion initiatives
do not determine political traditions, citizen values, the nature of major political
organizations, or the strength of anti-democratic groups. Since these are the po-
litical forces that shape the trajectory of new democracies most, the role of for-
eign assistance is rarely, if ever, determinant.
In sum, democracy in Timor-Leste was locally grown. It was nurtured by for-
eign actors, but foreign resources did not guarantee either its founding or its du-
rability. A whole host of national heroes had come to embrace democracy during
the war of resistance and had started building inclusive pluralist institutions on
their own—before international aid was forthcoming. This is why outsiders sur-
veyed the scene in 2002 and concluded that “no” local institutions posed a “se-
rious threat” to democracy (Smith 2004, 157). Oisín Tansey is right to remind
us that “all the significant parties . . . were unequivocally pro-democratic” before
foreign funded democracy-building began and that Timor-Leste’s transition was
largely a process driven by local actors in which even the UN eventually “became
a minor partner” (Tansey 2009, 68–69).
Conclusion
My explanation for the resilience of Timor-Leste’s democracy resonates with

many of the conclusions drawn from the other cases in this volume. We see
parallels with South Africa, and Benin regarding the decisive role of leaders’
perceptions: in Timor-Leste, as in both of these very different cases, democracy
emerged and endured, in part, because all of the major political actors had
“faith” in their ability to survive under democratic rule. This faith derived
from an extreme version of the “authoritarian weakness” that helps explain
democracy’s resilience in Benin, Ukraine, and elsewhere: resistance leaders in
Timor-Leste rightly recognized the profound weakness of the institutions cre-
ated by the occupation forces and thus had faith that democracy would work to
their advantage.
The decisive role played by resistance leaders offers striking parallels with the
case of India. In India, as in Timor-Leste, it was emerging and not incumbent
elites who mattered most because those who eschewed participation in the inde-
pendence movement simply could not “be counted as powerful political actors.”
The “democratic temper” that these emerging elites developed during decades of
struggle proved decisive in both cases, as well.
The subject of “democratic temper” brings us to a parallel between Timor-
Leste and cases as diverse as Argentina and Indonesia. We see parallels with
Argentina in the discrediting of extremism and in the pivotal role of changes in
elite “normative preferences.” And we see parallels with Indonesia not simply in
the inclusive/egalitarian nationalism adopted by key elites but in the very mech-
anism through which adoption occurred. In Indonesia, as in Timor-Leste, the
inclusive form of nationalism that eventually worked to “secure” and “sustain”
its 1999 democracy was in part “a product of its revolutionary path to indepen-
dence” (Chapter 3, this volume). In Timor-Leste, Indonesia, India, and much of
Southern Africa, durable democracy emerged not from slow, structural changes
but from “extreme political, economic, and ideological crisis” instead (Chapter 4,
this volume).
Despite our longing for parsimonious theory, no outcome as important as
democratic durability can be attributed to a single explanatory variable, even in a
single case. The puzzling longevity of democracy in Timor-Leste is no exception.
But durability can be explained through generalizable causal stories, and I have
offered one here. In this case (and in many others), the roots of democratic du-
rability lie in conflict-history, that is, in how ideas and institutions changed (or
failed to change) during the conflicts that preceded democratization. My argu-
ment, like many of the others in this volume, is about the impact of what Slater
calls “historical inheritance” (Chapter 3, this volume). As I stated at the outset,
armed conflict per se does not boost democratic longevity: only certain sorts of
conflict-histories have positive effects. New democracies emerging from revolu-
tionary conflicts and new democracies emerging from successful separatist wars
are associated with increased durability. It is essential to underscore the fact that
separatist conflict is only helpful if it produces a new democracy in a new state.
Most separatist conflicts do not produce separate states, and even fewer become
democracies when they do.
Timor-Leste’s new democracy was disadvantaged by a whole host of structural

factors, but because it emerged on the heels of a successful separatist war (with
revolutionary beginnings), it was advantaged as well. As the historical analysis
presented here illustrates, Timor-Leste’s conflict-history left four legacies that
undergirded democratic survival. First, it weakened the ranks of the elites and
parties that had formed the local anti-democratic coalition. Second, it strength-
ened a potent and inclusive nationalism that not only replaced the diffuse identi-
ties of the past but gave rise to a series of inclusive institutions. Third, it bolstered
a competitive party system with genuinely popular leaders who could win free
elections. Finally, it produced a professionalized, pro-democratic military.
Most of these changes were wrought during twenty-four years of separatist
war. Yet the country’s first internal war left prodemocratic legacies as well. Like
many revolutionary struggles, it projected a framing of “the people” and “a na-
tion” that was ethnically and religiously inclusive, produced a number of elect-
able heroic figures, strengthened an electable party, and laid the early foundations
for a military apparatus that obeyed civilian authority.
***
I am not alone in relating regime durability to particular types of nationalism

(Tudor 2013; Tudor and Slater 2016; Varshney 2015), to pro-democratic leaders
(Mainwaring and Pérez-Liñán 2013), to strong parties (Riedl 2014), or to sup-
portive militaries (Brownlee et al. 2013; Levitsky and Way 2013; Slater 2020).
Each of the collaborators in this project has related regime durability to at least
one of these factors, and most have done so in studies of democratic regimes. My
findings here contribute to the literature by showing how variables used by others
in very different cases help explain both a failed and a successful attempt at de-
mocratization in Timor-Leste. But the chapter makes other contributions as well.
One contribution lies in advancing our thinking about how the factors that
bolster regime longevity actually emerge. Here, too, my collaborators have
made great strides. They have urged us to recognize the lasting importance of
“inherited political structures” (Brownlee et al. 2013, 43), to see that “strong
coercive and party organizations” are the fruit of long-term processes and not
“short-term crafting” (Levitsky and Way 2010, 83), to recognize that “divergent
pathways to democracy” have considerable “long run” consequences (Riedl
2016, 124), and to see the strong connection between revolutionary histories
and the parties and militaries that sustain dictatorships (Lachapelle et al. 2020;
Levitsky and Way 2013). I extend this literature by linking democratic durability
to the ideational and political structures inherited from conflicts of all sorts, in-
cluding revolutionary and non-revolutionary struggles alike. In so doing, I ex-
plore and highlight a cause of the causes we know to be important.
The study’s third contribution emanates from expanding our thinking about
democracy to literally new (as opposed to merely different) states. Our most in-
fluential arguments about the evolution of durable democracies have emerged
from cases where a new democracy replaces a dictatorship in an already existing
state (Haggard and Kaufman 2016; Linz and Stepan 1996). These arguments
focus on incumbent parties, on incumbent authoritarian elites, and on the co-
ercive apparatus of existing states. What happens if the regime and the state are
new? Studying democracies emerging from successful separatist wars gives us a
more refined understanding of the varieties of identity conflict and better pur-
chase on the critical policy questions they generate.
The final contribution of this chapter lies in offering an explanation for de-
mocracy in hard places. The institutions and ideas that are helpful to demo-
cratic longevity have often emerged from the cluster of slow-moving changes
wrought by economic development. I have shown that these same factors can
also be brought about by the disequilibria and dilemmas caused by certain types
of war. This argument is certainly counterintuitive, and by associating something
evil with something good it is normatively problematic. Yet it is also perversely
hopeful—for it shows us that even poor countries can enjoy democracy if they
are mobilized to fight for it.
Notes
* The author thanks Mário Rebelo for excellent research assistance.

1. External war can have positive effects on democratic durability, too (Bermeo 2010).
2. I thank Steve Fish, Rachel Riedl, and the participants in this project for this and other
helpful suggestions.
3. By new democracies I mean regimes that are given a Polity score of 6 or above in
Marshall, Gurr, and Jaggers (2018). My results are robust to using the democracy
coding of Boix, Miller, and Rosato (Boix-Miller-Rosato [BMR] dichotomous coding of
democracy, 1800–2015). I emphasize my exclusive focus on “new democracies.” I make
no generalizations about the effects of war on the likelihood of democratic transitions.
(Indeed, the vast majority of wars are not followed by new democracies.) I also make
no generalization about war and the durability of new (or existing) dictatorships. For
insights on war and dictatorship, see Lachapelle et al. (2020). For more discussion of
the effects of conflict history on democratic durability, see Bermeo (forthcoming).
4. Secessionist rebels are often advantaged in several ways by their wartime conduct.
When they try to build strong local support and get recognition and resources from
the international community, they are less likely to kill civilians and more likely to pro-
vide civilians with services. See Jo 2015; Mampilly 2011; and Staniland 2012.
5. Gusmão enjoyed a landslide win in the 2002 presidential elections as an independent.
6. Timor-Leste’s post-independence state institutions were affected by what scholars have
aptly termed an “anti-past” vision—that is, “an inversion of everything wrong” about
the Indonesian regime (Bovensiepen and Nygaard-Christensen 2018, 423), including

military rule.
7. This was discovered before independence and is a major reason Australia turned
a blind eye to the Indonesian invasion. Australia and Timor-Leste were involved
in a costly maritime boundary dispute until 2018 which hampered optimal use of
the resource revenue too. For a brief overview with a discussion of democracy, see
Feijó (2018).
8. Roll shows how the pensions funded by oil money were key to both state-building and
stabilization. Like the oil wealth in Middle East dictatorships, oil revenue was some-
times used to “contain challenges” (Brownlee et al. 2013, 30).
9. Quite the contrary, it may, in the future, harm the legitimacy of elites who are already
being accused of oil-and gas-related corruption (Kingsbury 2018).
7
Economic Crises, Military Rebellions, and
Democratic Survival
Argentina, 1983–2021
Scott Mainwaring and Emilia Simison
What enables democracies to survive under difficult circumstances?* Why do

actors support democratic regimes when they experience deep material losses?
This chapter addresses these questions, focusing on the survival of democracy in
Argentina since 1983.
Argentina was a difficult case for building enduring democracy in different
ways from the other countries studied in this volume. At the time of the transi-
tion to democracy, Argentina was wealthier than the other countries. Nor did it
face the challenges of religious, linguistic, and ethnic diversity that Benin, India,
and Indonesia did; these cleavages are not very salient politically in Argentina.
The challenges in Argentina were rather a past history of repeated failures of de-
mocracy and of authoritarian actors that had repeatedly subverted democracy,
plus three severe economic crises and military rebellions.
Our argument unfolds in two stages. The first is that the military dictatorship
of 1976–83 was a dreadful failure. Its economic results were terrible; it waged a
disastrous war against the United Kingdom by invading the Falkland/Malvinas
islands in 1982; and it committed more human rights abuses than any previous
Argentine dictatorship in the twentieth century. These outcomes generated a
widespread repudiation of the past that led actors across the political spectrum
to a new valuing of democracy and to a rejection of extremism.
Since 1983, Argentina has been a world leader in memorializing the repudi-
ation of the dictatorship’s crimes and in creating a collective memory about the
importance of human rights. This effort has made it difficult for authoritarian ac-
tors to gain a political toehold and has helped sustain commitment to democracy.
The second stage of the argument focuses on two proximate explana-
tory variables: the demise of extremist anti-system actors and actors’ chan-
ging normative preferences about authoritarianism and democracy. Extremist
anti-system actors on the left and right bear great responsibility for the de-
struction of Argentina’s previous democratic experiment from 1973 to 1976.

Economic Crises, Military Rebellions, and Democratic Survival 189
The revolutionary left, right-wing death squads, and by 1975 the military en-
gaged in widespread violence that left much of Argentine society clamoring for
a coup that finally arrived on March 24, 1976. After the coup, through ruthless
repression, the dictatorship crushed the revolutionary left. What remained of
it rejected the totalitarian temptation and largely accepted democracy (Ollier
2009). The revolutionary left ceased to exist as a meaningful actor after 1983. The
extremist authoritarian right was also much weaker after 1983. The fiasco of the
dictatorship from 1976 to 1983 discredited the authoritarian right and the mili-
tary. Although the military remained a threat to the new regime at the time of the
democratic transition, it was a chastened actor. An important part of the story
of democratic survival after 1983 has to do with the demise of the revolutionary
left and the paramilitary death squads, the acceptance of democracy within the
armed forces, and the gradual establishment of civilian control over the mili-
tary. Without these extremist anti-system actors, democracy stood a much better
chance.
The second proximate factor that helped democracy survive after 1983 was
that the actors that remained important after 1983 shifted their attitudes about
democracy. If political parties, other organized actors, and citizens value dem-
ocratic rights and procedures, a regime can survive severe crises. Conversely, as
Argentina’s breakdowns of 1951,1 1962, 1966, and 1976 showed, if most powerful
actors are indifferent toward democracy or prefer an authoritarian regime, de-
mocracy can crumble under economic conditions that were much better than
those from 1983 to 1990, 1998 to 2002, and 2012 to 2020.
Graphically, the argument has the form presented in Figure 7.1.
In Argentina, democracy broke down in 1930, 1951, 1962, 1966, and 1976 de-
spite the country’s relatively high income and moderate inequalities. It survived
Vastly weaker set of actors

Demise of the revolutionary with a normative
A disastrous military preference for dictatorship
left and discrediting of
dictatorship and radical policy
military and extreme right
preferences
Building a human
rights memory
Stronger core of actors with

International Revaluing of democracy a normative preference for
influences by other actors democracy and with
moderate policy preferences
Figure 7.1. Graphic summary of the argument

after 1983 despite hyperinflation, three severe recessions, a dramatic worsening

of poverty and income inequality, and several military rebellions. But the actors
and their attitudes about democracy changed radically in Argentina’s democratic
experiment that began in 1983, and this is the key to understanding why democ-
racy survived.
Why Do Actors Support Democratic Regimes?
The survival of democracy hinges on the balance of power between forces

that support or accept the regime and those that would like to undermine it.
Democratic regimes survive when powerful actors support them. They collapse
when the actors that want to undermine them are more powerful than those that
support democracy.2 Why, then, do actors support democracies?
An instrumental material logic to this question focuses on economic
benefits: actors support a democratic regime if it generates good material
outcomes for them. Some class approaches that focus on whether different
classes gain materially exemplify this instrumental material logic. Acemoglu and
Robinson (2006) and Boix (2003) claim that the poor support democracy when
revolution is not feasible because democracy redistributes income to the poor.
The wealthy support authoritarianism under conditions of high inequality be-
cause democracy redistributes income to the poor, and they want to avoid this
outcome. These class and some other political economy approaches (Ansell and
Samuels 2014) focus exclusively on material gains and losses.
Support for regimes is undoubtedly sometimes conditional on material
outcomes. If a new democracy produces poor material outcomes over an ex-
tended time, some actors are likely to withdraw their support. Yet countless dem-
ocracies, including Argentina in the 1980s, from 1998 to 2002, and again since
2012, have survived bad economic and social performance (Remmer 1996). We
therefore layer four additional assumptions on top of the instrumental material
explanation for supporting regimes.
First, some actors support or oppose a regime based on other kinds of policy
benefits. Individuals and organizations might support a regime because of ed-
ucation policies, cultural and social policies (abortion, gay rights, etc.), immi-
gration policy, or policies around religion, language, ethnicity, and nationalism.
For example, conservative Islamic actors in Iran support the regime primarily
because of its religious and moral policies. However, even if we include other
policy outcomes, many democracies have performed dismally and still survived.
In Latin America, many democracies have survived despite high crime rates and
corruption scandals in addition to extended periods of poor economic and so-
cial results.
Second, actors and individuals implicitly compare their expected utility under
democracy with their expectations about utility under dictatorship. They might
believe that a democratic regime has produced poor substantive outcomes, but
they lack confidence that a dictatorship would do better. In that case, they would
have no incentive to support a move to dictatorship. They might not actively sup-
port the democracy, but neither they would actively support a coup. This per-
spective that actors implicitly compare utility under democracy and dictatorship
helps illuminate the logic of some Argentine actors after 1983. Real wages fell
dramatically in Argentina from 1983 to 1990 (by 30 percent) and from 1998 to
2002 (by 27 percent) (ECLAC 1993, 2004). However, the recent memory of wage
contraction under the military dictatorship reduced workers’ willingness to sup-
port another coup.
Nevertheless, an approach that posits that actors focus exclusively on material
outcomes and compare utility under different regime types often has difficulty
explaining actors’ positions. Under Argentina’s semi-democratic regime of 1963
to 1966, the economy grew at a rate that had few matches in the twentieth century
(Gerchunoff and Llach 2010, 300). Per capita GDP grew by 8.5 percent in 1964
and 8.9 percent in 1965 (but then fell 2.1 percent in 1966) (World Development
Indicators, n.d.).3 Growth was much better than it had been during the two pre-
vious authoritarian regimes (1955–58 and 1962–63). Unemployment dropped
from 8.8 percent in 1963 to 5.2 percent in 1966 (García Vázquez 1994, 294), and
real wages grew more than 10 percent (Gerchunoff and Llach 2010, 296). Yet or-
ganized labor and most business groups supported the 1966 coup that overthrew
the government. In light of the weak economic performance of the previous
dictatorships (1955–58 and 1962–63), it is unlikely that most actors concluded
that a new dictatorship would have better economic performance.
Third, actors might incur some costs for opposing democracy, or they might
be constrained by strong institutions. Most individuals and organizations would
consider the probability of a successful attempt to overthrow a democracy and
the likely costs of attempting but failing to do so. Even if opposition actors de-
spise the incumbent democracy, if the probability of overthrowing it is very low
and the potential costs of supporting a failed attempt to achieve a regime change
are high, most actors will remain on the sidelines.
However, if a democracy produces bad substantive outcomes for most pow-
erful actors, the costs of signaling subtle support to change the regime should not
be terribly high. Under democracy, actors can signal opposition to the regime at
a much lower cost than they can under repressive dictatorships. Because liberal
democracies are reluctant to repress anti-system actors that do not engage in vi-
olence or otherwise go beyond the law, these actors have considerable latitude to
signal their support for a regime change. If actors choose regimes based on what
they deliver substantively, under circumstances of extended poor performance,
a broad coalition of actors should oppose it. Powerful opposition to a democracy

does not ensure that it will erode or break down, but it makes this outcome much
more likely. If a regime produced bad outcomes for vast swaths of society, as has
been the case in Argentina for extended times since 1983, and if actors supported
or opposed regimes based only on material outcomes, it would be difficult for
democracy to survive.
Fourth, some individuals and organizations have preferences about the re-
gime in addition to caring about substantive outcomes. By definition, democ-
racy affords free and fair elections, nearly universal adult voting and citizenship
rights, respect for a panoply of political and civil rights, and civilian control over
the military. Many citizens, political leaders, and organized actors might value
these procedural aspects of democracy. They might not benefit materially from
democracy, and they might endure policy losses. However, following Winston
Churchill’s famous maxim that “democracy is the worst form of government ex-
cept all those other forms that have been tried,” they still might conclude that
democracy is the best possible regime. Voters emerging from the tyranny of
communist rule valued democracy even when most post-communist econo-
mies were struggling in the 1990s (Hofferbert and Klingemann 1999; Rose and
Mishler 1996). Following Mainwaring and Pérez-Liñán (2013), we refer to actors
that value the procedural aspects of democracy as having a normative prefer-
ence for democracy. We argue that in Argentina, a change in actors’ normative
preferences and the demise of extremist anti-system actors are key to explaining
the survival of democracy after 1983.
We divide actors’ utility into two broad baskets. The first basket includes all
outcomes and policies that are not related to the political regime. Second are the
outcomes and procedures that are an intrinsic part of democracy as a political
regime.4 Liberal democracy has four characteristics:
1. Free and fair elections for the national assembly and the head of
government.
2. Almost universal adult suffrage among citizens and no exclusions from cit-
izenship that were not common at a given historical moment.
3. A broad set of rights (freedom of expression, freedom of speech, freedom
of assembly, freedom of religion, freedom from prosecution based on po-
litical beliefs), and state institutions that function to ensure these rights.
4. Civilian control over the military and ability to choose policies without
having them dictated by armed actors.
Many individuals and organizations value these defining features of

democracy—as Ashutosh Varshney argues in his chapter on India. Human
rights organizations value rights that are an intrinsic part of democracy. In
democracies, most journalists, media companies, artists, and academics value

freedom of expression. Many churches, religious leaders, and followers value
freedom of religion. Many people believe in basic principles of human dignity
and sanctity; such beliefs could prompt favoring democracy and could lead
to moral repudiation of repressive authoritarian regimes. Some citizens want
the right to choose their leaders in free and fair elections. Political parties and
politicians might want to maintain free and fair elections as a hedge against au-
thoritarian regimes that cut them out entirely—or worse, that ban and perse-
cute them. Many individuals value procedural fairness and not only substantive
outcomes (Frey, Benz, and Stutzer 2004). Democracy is more likely than most
authoritarian regimes to offer procedural fairness in who gets to run for office,
how they get elected, who governs, and how they govern. Democracy is also
more likely to offer fair legal procedures. Some business leaders might value the
checks and balances and procedural guarantees of democracies. Some individ-
uals have a moral compass that leads them to repudiate the rights violations that
authoritarian regimes commit and to value democracy’s guarantees of rights.
When individuals or organizations give some weight to these considerations that
are defining features of democracy, they have a normative preference for democ-
racy. This preference is based on the guarantees, procedures, and rights that de-
mocracy enshrines.
Skeptics believe that actors favor democracy only out of self-interest. Indeed,
some powerful Argentine actors (the Peronist Party and the labor movement,
for example) whose attitudes about democracy underwent profound transform-
ations acknowledged after 1983 that democracy was in their interests. But these
actors did nothing to support democracy from 1973 to 1976, and they passively
accepted a coup in 1976. If democracy was not in their (perception of their)
interests between 1945 and 1976 but was in their (perception of their) interests
after 1983, it is because they came to a new valuing of democracy after the harsh
realities of the military dictatorship. Actors’ perceptions of their interests are
deeply infused with their normative beliefs.
Organizations and individuals weight instrumental policy interests and nor-
mative preferences about the regime differently. Many might weight substantive
outcomes other than those that are defining features of a democracy for almost
100 percent of their implicit calculus. But others value the procedures and rights
that democracy guarantees. This orientation might be especially prevalent after
individuals and organizations have experienced or witnessed deprivations be-
cause of dictatorships, or because they have been socialized in well-functioning
democratic regimes.
If political actors value the rights and procedures that are defining features
of democracy, democracy is more likely to survive. A normative preference for
democracy refers to actors’ willingness to incur other policy costs in order to
defend or achieve democracy; an actor values democracy because of these rights

and procedures. It does not mean that they always prefer democracy at any cost.
Normative preferences about democracy include some substantive outcomes
(protection of rights), and they can also be based on “procedural utility”—the
well-being derived from procedures above and beyond the outcomes they gen-
erate (Frey et al. 2004).5
When actors value the rights and procedures of democracy, it can survive
bad governing performance (Linz 1988; Linz and Stepan 1989; O’Donnell
1986; Remmer 1996). They are less willing to subvert democracy. Conversely,
in contexts of weak accountability institutions, democracy that rests exclusively
on regime performance is almost inevitably precarious. With weak accounta-
bility institutions such as courts and the national legislature, if no actors were
committed to democracy, powerful narrowly instrumental actors would not
feel constrained by the rules of the game. If actors preferred democracy exclu-
sively because of the instrumental advantages it confers to them, any equilibrium
would be vulnerable to easy disruption unless institutions such as courts and
the congress were strong, or unless a tight balance of power remained in place
over time.
Preferences are not static; individuals and organizations change how much
they value democracy on its intrinsic merits. The legacies of different types of
authoritarian regimes, for example, affect future support for democracy; people
who have experienced exclusionary authoritarian regimes are less likely to hold
anti-democratic beliefs (Neundorf et al. 2019). In fact, survey data show that
Argentines valued democracy more after 1983, almost certainly in response to
the gross human rights violations, ineptitude of the last military dictatorship,
and the powerful memorialization of human rights abuses.
The values that individuals and organizations implicitly assign to these two
categories of outcomes and procedures are not observable. The point is that some
individuals and organizations attach a positive value to the rights and procedures
that characterize democracy. This positive value is part of the glue that some-
times enables democracies to survive in difficult conditions.
No observational evidence can definitively show that actors support a de-
mocracy solely because of its other substantive outcomes or also in part because
they value the rights, guarantees, and procedures of a democracy. However, in
Argentina, large swaths of society experienced major economic losses from 1983
to 1990, 1998 to 2002, and 2012 to 2020. Despite these losses, powerful organ-
ized actors did not merely accept democracy because they had no choice; they
actively defended the regime in the face of military rebellions. Therefore, we can
definitively rule out simple material explanations.
Moreover, the discourse and behavior of powerful actors changed profoundly
relative to the past, and these actors actively supported democracy. A look at the
discourse and behavior of some actors suggests that they had a normative pref-
erence for democracy and that this preference helped democracy survive. The
demise of radical anti-system actors also made it easier for democracy to survive;
it is an equally important part of our account.
A Failed Dictatorship
The dictatorship of 1976–83 was a spectacular failure. Initially, it enjoyed broad

support despite massive human rights violations. However, this support dwin-
dled as the economy sputtered, and then it collapsed when the generals initiated
a war against the United Kingdom by invading the Falkland/Malvinas Islands.
The damage wrought by the dictatorship pushed important actors and countless
citizens toward a commitment to democracy.
Economic Results
The dictatorship’s economic policies failed dismally. As the Economic

Commission for Latin America and the Caribbean reported, when the new dem-
ocratic government took office in December 1983, “The economic problems
were very grave. . . . The per capita product was similar to what it had been
fifteen years earlier. . . . Domestic savings and the availability of external fi-
nancing had declined sharply; the government was operating at a great deficit;
workers . . . were trying to recoup levels of consumption that had been put off
for a long time . . . and external debtors were reluctant to refinance their credits
much less offer new loans. . . . The very high inflation rate revealed a large dise-
quilibrium in public finances . . . and stimulated speculative behavior in financial
and exchange rate markets” (ECLAC 1985, 1–2). Real wages fell by 41.3 percent
between 1975 and 1982 (Graña and Kennedy 2008, table 7). Per capita GDP was
lower in 1983 than it had been in 1975, the last full year of democracy (World
Development Indicators, n.d.). The external debt ballooned, leaving an enor-
mous burden for the new government. Inflation surged, reaching 380 percent
in 1983 (World Development Indicators, n.d.). The economic legacy for the in-
coming democratic regime was ruinous.
A Reckless and Disastrous War
By 1981, the dictatorship was in disarray, with support faltering because of the
economy mismanagement. Military cohesion had cracked, and public opinion
had turned sharply against the dictatorship. In response to the crumbling sup-
port, the dictatorship gambled that it could galvanize public support by invading
the Falkland/Malvinas islands. Both the British and the Argentines had claimed
the islands since the nineteenth century, but the British had long exercised
control.
Waging war against the United Kingdom was a reckless and disastrous
gamble. The dictators launched the war by invading the Falklands on April
2. Their expectation that the invasion would generate renewed public support
was absolutely right in the short term. Because the British had almost no military
presence on the islands, the Argentine troops quickly took over without blood-
shed, generating a jingoistic response that the dictatorship had expected. In the
UK, the Thatcher government responded with equal nationalistic fervor and de-
termination, quickly dispatching a fleet of warships across the Atlantic.
To garner support for the war, until the day the Argentine military surren-
dered, with a gigantic propaganda output, the generals proclaimed that they
were winning the war. The press coverage was massive and jingoistic; the tightly
controlled media proclaimed that the Argentine military was winning. Most
Argentines swallowed the lies and applauded the invasion; even much of the left,
which had been the object of savage repression by the dictatorship, rallied to the
cause. Millions of Argentines mobilized on the streets to express their support
for the war.
It took several weeks for British warships to reach the islands and engage the
Argentine military, initially in naval battles. The British military landed on East
Falkland Island on May 21 and subsequently routed the Argentines, leading to
the surrender on June 14, 1982. The cost in lives (about 649 Argentine soldiers
died and another 1,177 were wounded) and economic ruin was high. The lo-
gistical preparation was dismal, leading to appalling conditions for the mostly
young Argentine conscripts who invaded. Without adequate food, boots, and
coats, they starved and froze on the cold, wind-swept islands until the generals
capitulated. Argentine military documents that came to light in the 2010s in-
dicate that some officers tortured and abused the conscripts to keep them in
line.6 The dictatorship’s brazen lies about winning the war reinforced the effects
of the bellicose disaster, discrediting the military as a bunch of incompetents
and liars.
The fallout was immediate and profound; it brought down the dictatorship,
although the actual transfer of power did not take place for almost eighteen
months, on December 10, 1983. President Leopoldo Galtieri was forced to resign
only four days after the Argentine surrender, June 18, 1982, and he was placed
under (comfortable) house arrest in a military compound. His entire cabinet
resigned the same day, although they remained in place until new appointments
could take over.
Human Rights Abuses and Their Memorialization
The dictatorship of 1976–83 was Argentina’s most repressive of the twentieth

century. The country’s governments, courts, human rights NGOs, and countless
artists, scholars, and journalists have helped to build an enduring commitment
to never allowing something similar to occur. The memorialization of human
rights abuses under the dictatorship has helped build and preserve an anti-
authoritarian sentiment.
Few countries in the world have ever done as much as Argentina to me-
morialize the atrocities of the terrible past and to prosecute those respon-
sible for human rights crimes under a dictatorship (González Ocantos 2016,
2020; Sikkink 2008). The report commissioned by the Alfonsín government
in its first days in office, Nunca Más: Informe de la Comisión Nacional sobre
la Desaparición de Personas (Never Again: Report of the Argentine National
Commission on the Disappearance of Persons), has sold more than 700,000
copies, and it had a huge impact (Crenzel 2015). Argentina became a world pi-
oneer in prosecuting military dictators who presided over gross human rights
abuses. In 1985, the world’s most publicized human rights trials since the
Nuremburg trails in Germany in 1945–46, galvanized public opinion against
the atrocities committed by the military. The top leaders of the dictatorship
ended up in jail. President Carlos Menem (1989–99) later pardoned them,
but subsequent court cases put them back in jail. As of November 2018, 3,007
individuals had faced charges for human rights abuses during the dictatorship
(Human Rights Watch 2019), and by July 2019 courts had convicted 915 indi-
viduals of human rights abuses (González-Ocantos 2020, 14). Many trials and
convictions were highly publicized, creating an ongoing sense of a repudiation
of the terrible past.
In 2002, the Argentine Congress created the National Day of Memory for
Truth and Justice, commemorated as a national holiday every year on March 24,
the day of the 1976 coup. Around that day, all schools focus on human rights. The
country continues to have one of the most vibrant human rights movements in
the world.
Prize-winning films such as The Official Story (1985) and The Secret in Their
Eyes (2009),7 both of which dramatized the human rights abuses committed
under the dictatorship, drew huge audiences in Argentina and internationally.
Countless other films, works of theater, works by journalists,8 memoirs, novels,
and academic works also memorialized the gross human rights violations. As
of 2019, Argentina had thirty-four official public museums, parks, and other
institutions that “have permanent activities related to the transmission and pro-
motion of human rights.”9 Among the best known is the Museo Sitio de Memoria
ESMA (Museum and Site of Memory ESMA) in Buenos Aires, which opened in
2015. It functions at the site of what had been the largest torture and extermina-
tion center during the dictatorship.10
The reinforcement of the importance of human rights and democracy coming
both from state institutions and civil society organizations has reduced the space
for authoritarian candidates, parties, social movements, and civil society organi-
zations. Conversely, these efforts have kept alive a message about the importance
of human rights and democracy.
The trials against the former dictators in 1985 and the guilty verdicts in
December 1985 riveted the attention of the world’s human rights and democracy
advocates on Argentina. The courts effectively prosecuted many abusers, com-
municating a message about the importance of human rights (González Ocantos
2016). The governments of Néstor Kirchner (2003–7) and Cristina Fernández
de Kirchner (2007–15) reopened trials against the former abusers, invoked the
importance of human rights, and criticized the dictatorship’s abuses, activating
the memory of how depraved that dictatorship was and affirming the value of
democracy.
Advantages and Three Potential Threats to Democracy
Argentina had five democratic advantages in 1983—three domestic and two

international. First, it was an upper-middle-income country with one of the
highest standards of living and highest educational levels in Latin America.
Second, although income inequality had worsened during the previous
decade, it remained one of the least unequal countries in Latin America.
Third, the dictatorship of 1976–83 was thoroughly defeated and discredited
after its disastrous invasion of the Falkland/Malvinas Islands in 1982, and
after its mishandling of economic policy. The discrediting of the military
gave operating space to the new democratic government. Fourth, the neigh-
borhood and region were starting to transform in a democratic direction.
After several false starts and coups between 1978 and 1982, Bolivia com-
pleted a precarious transition to democracy in 1982. Brazil’s and Uruguay’s
transitions to democracy were well under way. These changes were incom-
plete when Argentina transitioned back to democracy in December 1983, but
the tide was shifting away from dictatorship. A democratic neighborhood
makes it easier for a country to sustain democracy. Fifth, notwithstanding
considerable wavering by the Reagan administration in Central America, US
policy had shifted toward putting more emphasis on supporting democracy
in South America. In the Argentine case, the US veered toward supporting
a democratic transition after the dictatorship initiated war with the United
Kingdom in 1982.
Given these advantages, Argentina was not among the most difficult cases
of democracy, but it was not an easy case from 1983 through the early 1990s.
Democrats in Argentina faced three formidable threats to democracy: a long an-
tecedent history of coups and authoritarian actors, severe economic crises, and
military rebellions.
Argentina had a long history of democratic failures despite a high level of mod-
ernization and moderate inequalities, so its structural advantages did not guar-
antee democratic survival. The country had experienced democratic breakdowns
in 1930, 1951, 1962, 1966, and 1976. It was one of the world champions of
breakdowns in the twentieth century. Powerful authoritarian actors had repeat-
edly subverted democracy. Previous studies have shown that a past history of
breakdowns is a good predictor of a future breakdown (Przeworski et al. 2000,
124, table 2.17). Given this fact and the reality that no previous experience with
democracy in Argentina had begun under economic conditions nearly as bad as
those that the Alfonsín government faced, and that the country was more un-
equal and marginally poorer than it had been at the beginning of the previous
failed democratic experiment (1973–76), democratic survival was not a foregone
conclusion.
Argentina has confronted three punishing economic crises, with three of the
largest declines in per capita income under democracy in Latin America since
1978. Only the Peruvian per capita GDP decline from 1980 to 1992 was greater
than the declines Argentina experienced from 1983–90, 1998–2002, and 2012–
20 (World Development Indicators, n.d.),11 and democracy in Peru broke down
at the end of that period, in April 1992. By 1990, Argentina’s per capita GDP
was 17.5 percent lower than it had been in 1984 (World Development Indicators,
n.d.),12 and inflation reached 4,923 percent in 1989 (CEPALSTAT, n.d.).13 Per
capita GDP dropped 22.0 percent from 1998 to 2002, accompanied by a massive
increase in unemployment and poverty. Another long and cumulatively terrible
slump began in 2012 and culminated in an estimated 10.8 percent decline in per
capita GDP in Covid-plagued 2020 (World Development Indicators, n.d.).14 By
2020, per capita GDP was 20.1 percent lower than it had been in 2011. No third-
wave democracy has survived three equally devastating economic crises.
These economic crises could have derailed democracy—especially the first
one, when the military leadership from the previous dictatorship was still partly
intact. Indeed, simple materialist understandings of what makes democracy
viable probably would have predicted breakdown. The second crisis provoked
massive social mobilizations. But at no time did these crises break democracy.
Whereas Argentina’s economic problems of 1929–30, 1966, and 1975–76 solid-
ified support for coups, the severe crises of 1989–90, 2001–2, and 2018–20 did
not. Focusing on a democracy that survived severe and prolonged economic
crises that led to losses for most sectors of society rules out simple materialistic
explanations for regime support. During the crises, most sectors of society expe-
rienced economic losses.
For the first part of this democratic period, roughly 1983 to 1990, Argentina
has one other advantage for case selection. Several military rebellions in the
1980s could have led to a breakdown, especially in tandem with the economic
crises. The investigations of human rights abuses committed by the dictatorship
proved to be destabilizing. At that time, the military was still divided between
factions that accepted civilian control and democracy and sectors that remained
wedded to the authoritarian tradition. Members of the authoritarian factions en-
gaged in uprisings in 1987 and 1988 that could have threatened the stability of
the democratic regime. However, the revolts failed to garner political support.
Large numbers of people, as well as the main political and societal actors, actively
mobilized in defense of democracy.
The frequent military rebellions distinguish Argentina from other transition
cases at that time. In McGuire’s (1995a, 194–95) analysis of seven transitions
from authoritarian rule in Latin America and Southern Europe, Argentina, with
four rebellions, had the most military uprisings. Therefore, it would be wrong
to argue that actors tolerated democracy only because they had no chance
of forming a winning coalition to install an authoritarian regime. Indeed, the
rebellions failed in part because they generated a massive outpouring of public
support for democracy and because powerful actors such as labor unions and
business associations came to the regime’s defense.
The notion that Argentina’s democratization was easy also ignores the
perspectives of actors at that time. Around 1982–84, many Argentine analysts
underscored the difficult challenges to democratic survival. The obstacles in-
cluded profound economic problems such as a massive foreign debt incurred
by the dictatorship, a history of an authoritarian political culture, a military that
although defeated was still profoundly authoritarian and powerful, a history of
many failed regimes, and weak democratic institutions. Acuña (1984, 261–62)
wrote that if business interests felt attacked by the economic policies of the new
democratic government, they might try to provoke a new military intervention.
Amadeo (1984, 94) doubted that a return to democracy would take place in the
medium term unless some deep structural problems were resolved. Bruno (1984,
128) warned that “actors that represented the oligarchy and imperialism” would
attempt to undermine democracy if the constitutional government did not pro-
tect their interests.
Cavarozzi (1983, 69–70) wrote that “from the perspective of the stabiliza-
tion of a democracy, the situation is not too promising.”15 De Riz argued, “The
democratic reconstitution of the relationships between the state and society in
Argentina does not seem to be attainable in the short term” (De Riz 1982, 1215–
16). Oszlak (1984, 12) wrote that the establishment of democratic practices and
institutions “will be a difficult and slow process.” O’Donnell (1983) warned that
“once again, the political opening could be perceived as a transitory refugee of
an authoritarianism in crisis, without a path to democracy.” Nearly a quarter-
century after the 1983 transition, well-known historian Luis Alberto Romero
(2006, 29) wrote, “It is unquestionably possible that without the military dic-
tatorship, democracy would have quickly sunk . . . because in 1983, there was
nothing great with which to build democracy, except for hope.”
Moreover, Argentine citizens were also skeptical that democracy would last.
In a national survey carried out in late 1985, 95 percent of respondents said that
“the authoritarian regimes will return if the government does not organize things
well,” against only 5 percent who believed that “democracy will become stronger
in our country, and authoritarian governments are something of the past” (CID
1985).16
Democracy in Argentina, 1983–2020
The new democratic regime began on December 10, 1983, when Raúl Alfonsín
assumed the presidency. Democracy in Argentina has now survived for thirty-
eight years, and it has had a solid quality. According to V-Dem and Freedom
House, since 1984, Argentina has consistently had one of the highest levels of
democracy in Latin America. Figure 7.2 shows Argentina’s V-Dem Liberal
Democracy Index scores and Freedom House scores from 1983 to 2020. Freedom
House scores are inverted so that a high score represents a high level of democ-
racy, and then standardized from 0 to 1 so that they are on the same 0-to-1 scale
as V-Dem’s Liberal Democracy score.17
Three Severe Economic Crises
Many scholars have shown that economic crises can imperil new democracies
(Gasiorowski 1995; Przeworski et al. 2000; Sing 2010; Svolik 2015). Economic
slumps that were much milder than what Argentina experienced in the 1980s
and 1998–2002 contributed to breakdowns in 1930, 1951/55, and 1976.
The dictatorship (1976–83) left the economy in a ruinous state, and the
Alfonsín administration never got a solid grip on the economy. From 1983 to
1991, the country lurched from one crisis to the next, culminating in the hyper-
inflation of 1989–90, when inflation reached 4,923 percent and 1,343 percent, re-
spectively (CEPALSTAT, n.d.). Few democracies have survived worse economic
conditions. Inflation remained high during Alfonsín’s whole term, ranging from
82 percent in 1986 to 4,923 percent in 1989 (CEPALSTAT, n.d.). GDP per capita
1.00
0.90
0.80
0.70
Democracy score
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
V-Dem score FH score
Figure 7.2. Argentina, V-Dem liberal democracy and Freedom House scores, 1983–2020
Sources: Coppedge et al. 2021; Freedom House 2021a.
diminished on average 2.1 percent yearly between 1981 and 1990, and the for-
eign debt reached 70 percent of the GDP (Novaro 2010, 226). According to the
IMF database on systemic banking crises (Valencia and Laeven 2012), the output
lost during the 1989–91 crisis reached 12.6 percent of GDP and the fiscal costs of
the crisis, defined as the component of gross fiscal outlays related to the restruc-
turing of the financial sector, were equivalent to 6 percent of GDP.
This crisis impoverished millions of Argentines and generated widespread
suffering and uncertainty. It precipitated the early resignation of President
Alfonsín on July 8, 1989, five months before his term was due to end, and the
early assumption of the presidency by his elected successor, Carlos Menem.
Despite the severity of the economic crisis, democracy survived. Argentina be-
came an early exemplar of how to displace a failed government but preserve a
democratic regime.
Another potential challenge to democracy occurred during the severe finan-
cial and economic crisis of 2001–2. The urban poverty rate hit 45.4 percent in
2002, shocking for a country where it had been only 16.1 percent eight years
earlier (ECLAC 2006, 317). Since ECLAC started regularly publishing data on
poverty at the beginning of the 1990s, no other country in Latin America regis-
tered such an extraordinary increase in poverty until Venezuela in the 2010s. We
wonder if there is another case of a democracy that experienced such an extraor-
dinary increase in poverty without a breakdown. During the crisis of 1999–2002,
per capita GDP declined sharply, from $8,729 to $6,817 (World Development
Indicators, n.d.).18 Unemployment surged from 7.4 percent in 1990 to 11.5 per-
cent in 1994, 14.9 percent in 1997, and 15.1 percent in 2000 (ECLAC, 2001).
By 1998, the economy had slid into a recession. With support from other
parties in addition to the Radical Civic Union (Unión Cívica Radical, UCR),
Fernando de la Rúa won the 1999 presidential election, in part because the high
economic and social toll had tarnished the Justicialist Party (Partido Justicialista,
PJ). The situation worsened in November 2001 with the collapse of the capital
market. In the face of massive demonstrations, De la Rúa resigned on December
20, 2001, beginning a period of great political instability and unpredictability.
Argentina had five different presidents in thirteen days, from December 20,
2001, to January 2, 2002. In Argentina’s past democratic periods, when actors
readily conspired against the regime, this instability could easily have led to a
breakdown. After resignations by four presidents, Congress elected Eduardo
Duhalde to serve from January 2, 2002, to May 25, 2003. By the time Duhalde left
office, the economic situation had improved greatly.
The economic and social costs of the 1994–2002 period were extremely high.
According to the IMF database on systemic banking crises, the output lost in the
2001–3 crisis was 70.97 percent of GDP, and its fiscal costs amounted to 9.6 per-
cent of GDP (Valencia and Laeven 2012). In terms of GDP drop, this crisis was
the worst by a Latin American democracy that was established in the third wave
and survived. According to the data of the Economic Commission on Latin
America and the Caribbean, no democracy in Latin America has experienced as
great an increase in poverty. Average incomes fell 22 percent between 1998 and
2002 (Steinberg 2015, 2). The national unemployment rate reached 19.7 percent
in 2002 according to the World Development Indicators, while the urban un-
employment rate was also 19.7 percent according to ECLAC (2004). The crisis
implied forty-two months of decline in the level of economic activity (Novaro,
Bonvecchi, and Cherny 2014, 25).
The third economic slump began in 2012 (−2.1 percent per capita growth),
followed by six years of cycling between low growth and recessions that con-
sistently more than wiped out the previous year’s tepid growth: 1.3 percent per
capita growth in 2013, a 3.5 percent drop in 2014, 1.7 percent growth in 2015, a
3.1 percent drop in 2016, and 1.8 percent growth in 2017. In 2018, the economy
began a steady decline with a drop in per capita GDP of 3.5 percent that year,
another decline of 3.0 percent in 2019, and a sharp drop of 11.5 percent in 2020
(ECLAC 2021, 117, table A1.3).
The Military Rebellions of the 1980s
Another challenge that Argentine democracy confronted was subordinating the

military. This was not an easy task. Although the dictatorship suffered a humili-
ating defeat in the 1982 war with the United Kingdom, the unreconstructed mil-
itary was still intact at the time of the democratic transition. Most of the military
had a profoundly authoritarian worldview (Franco 2009). Massive human rights
violations committed by the dictatorship created a delicate balancing act for the
government. On one hand, Alfonsín won the 1983 election in part because of his
firm stance on behalf of human rights. Human rights movements that had been
important actors in the struggle against the dictatorship demanded legal action
against the killers and torturers in the military and security forces. On the other
hand, the military and security forces had a high level of self-justification of their
activities during the “dirty war.” They believed that they were fighting to preserve
Western Christian civilization against a subversive threat. Acting against a pow-
erful entrenched military with this profile could have fostered a coup.
Of the new democracies in Latin America in the third wave, the Alfonsín
government clashed with the military more than any other, and it made more
efforts than any other foundational democratic government to punish human
rights abusers.19 As mentioned earlier in this chapter, just five days after his in-
auguration, Alfonsín formed a commission to investigate human rights abuses
during the dictatorship. The National Commission on the Disappearance of
People’s report, Nunca Más, documented the “disappearance” of 8,961 people

and the widespread use of torture and clandestine detention centers. It quickly
became a bestseller in Argentina and internationally, and it provoked interna-
tional outrage about the widespread human rights abuses. The Alfonsín govern-
ment prosecuted the leaders of the military dictatorship, and courts convicted
and sentenced them to lengthy prison terms in 1986.
In the context of a partly unreformed military, the ongoing investigations
into human rights abuses threatened to be destabilizing. The military was di-
vided between sectors that accepted civilian control and democracy and others
that remained wedded to the authoritarian tradition and engaged in uprisings
in April 1987 and January and December 1988 (Jaunarena 2011; Norden 1996).
During these military rebellions, citizens risked their lives by mobilizing on
the streets on behalf of democracy, and the main parties, labor unions, social
movements, and business groups came to the defense of democracy. This sup-
port helped enable democracy to survive the rebellions, and it was telling of the
shift in Argentine public opinion regarding military rule.
A new army revolt took place in late 1990, but loyal army units repressed
the insurgents. By the early 1990s, the military was under civilian control and
began to acknowledge excesses during the “dirty war.” Military extremism was
buried. The military declined in relevance after 1990 and ceased being an impor-
tant player in regime politics (Diamint 2006; Norden 1996; Palermo and Novaro
1996, 252). Since the early 1990s, there has been no chance of a successful mili-
tary coup.
Explaining Democratic Survival in Difficult

Economic Conditions
Why did powerful actors not mobilize against democracy during the severe ec-
onomic crises since 1983? Most powerful actors have been committed to pre-
serving democracy, and since the subordination of the military in the early 1990s,
there have been no important authoritarian actors. Normative commitments to
democracy helped the regime to survive economic crises that were more severe
than the economic problems faced during earlier breakdowns. Because all im-
portant actors have agreed that preserving democracy is important, the pro-
democracy coalition has remained intact even in moments of enormous turmoil
and economic crisis. All actors, even those that do not have a strong normative
preference, have accepted democracy.
The rest of this section shows how the preferences of the main political ac-
tors changed after the last military dictatorship. To trace these changes, we
used primary and secondary sources. Among the primary sources consulted
are the interviews available at the Oral History Archive of the Gino Germani
Institute,20 the Archive on Argentine Unionism at Torcuato Di Tella University,21
the General Confederation of Labor (Confederación General del Trabajo, CGT)
archives,22 and the digital archives of the newspapers Clarín and La Nación
(available since 1995).
The Justicialist (Peronist) Party
For democracy to endure, the preferences and behavior of governing parties and
presidents are especially important because they are the main actors that can
lead executive suffocations of democracy. Radical leftist governments can also
jeopardize democracy by promoting a backlash among conservative actors, po-
tentially leading to a military coup.
From 1983 to 2015, the Radicals and Peronists were consistently the core of
the democratic coalition (although the Radicals became a much weaker electoral
contender after 2001). The Peronist party, Partido Justicialista (PJ), experienced
a deep change in its attitudes toward democracy. Although Juan Perón was dem-
ocratically elected in 1945, he governed as an authoritarian populist until he was
ousted by a coup in 1955. From 1973 to 1976, Perón and his hand-picked pred-
ecessor and successor after his death in 1974 displayed indifference toward the
preservation of democracy. Most Peronist labor unions and other Peronist ac-
tors also did. The Peronist right, which included the country’s most powerful
right-wing death squads, and the Peronist left actively conspired against democ-
racy. An unreconstructed Peronism would have been disastrous for democracy
after 1983.
The Peronists won the presidency in 1989, 1995, 2003, 2007, 2011, and 2019,
and they were the main opposition party in 1983–89, 1999–2001, and 2015–19.
They left behind their more authoritarian tendencies. The highly authoritarian
revolutionary Peronist left and the right-wing Peronist death squads ceased
to exist. The disastrous experience of the dictatorship of 1976–83 led most PJ
leaders to value democracy. By 1984 the Renovación (“Renewal”) movement had
become the dominant force inside the PJ, pushing for further democratization
within the party. This movement included among its principles the true accept-
ance of party competition, faith in the Constitution, and the distinction be-
tween the partisan and corporatist spheres (Novaro 1994, 60–61). In the words
of one of Peronism’s most important leaders from the 1950s until 2005, Antonio
Cafiero, “It is necessary to learn to live together in a pluralist and fragile democ-
racy” (quoted in Novaro 1994, 61). This idea is clear also in declarations made by
Juan Gabriel Labaké, a member of the Supreme Peronist Command (Comando
Superior Justicialista), in 1985 on the steely determination of party members to
fight for the stability of the institutions and the constitutional order.23 A leading
Peronist politician and well-known political scientist and public intellectual,
Juan Abal Medina, wrote that “[t]‌he extremely difficult experience of the dicta-
torship led actors to recognize the value of . . . tolerance and diversity, which had
been absent in all of the previous democratic episodes” (Abal Medina and Suárez
Cao 2003, 78).24
Sectors within Peronism have been willing to sacrifice some democratic
practices (de Ipola 1987), but the movement distanced itself from the au-
thoritarian proclivities of its past. The PJ accepted its role as a loyal opposi-
tion (in Linz’s terms [1978b]) under democracy from 1983–89, 1999–2001
and 2015–19.
Once labor and the PJ became committed to democracy as a value, the
withering economic crises under democracy did not cause them to favor an
authoritarian outcome. In the midst of the 2001–2 financial meltdown, the
Peronist members of the Chamber of Deputies voted to approve March 24
(the anniversary of the 1976 coup) a day of memory of truth and justice: “The
Peronist deputies support this measure not only as Peronists but also as
democrats. . . . We have a commitment to democracy. Having suffered through
the infamous dictatorship, we must always celebrate the memory of those who
were disappeared, persecuted, and jailed, of our colleagues who needed to
leave the country, and of all those who fought to restore democracy and to en-
sure that we would never again have a dictatorship in our beloved country” (in
García Lerena 2007, 613).25
The military revolt of April 1987 showed the profound transformation among
PJ leaders, who offered unwavering support to the Radical administration.
Notwithstanding the failures of Alfonsín’s economic policies, the PJ did not at-
tempt to destabilize the regime. During the military revolts, the transformation
among partisan actors was clear. As Menem expressed in January 1989, the dem-
ocratic regime was not going to be jeopardized (quoted in Masi 2014, 319).
The five Peronist presidents who lasted more than a few weeks (Carlos
Menem, 1989–99; Eduardo Duhalde, 2002–3; Néstor Kirchner, 2003–7; Cristina
Fernández, 2007–15; Alberto Fernández, 2019–) were much more respectful of
democratic norms than Juan Perón had been from 1946 to 1955 and from 1973
to 1974, and than Isabel Perón had been from 1974 to 1976. They occasionally
ran roughshod over mechanisms of intrastate accountability, but they trespassed
within bounds that did not undermine democracy even when congressional
majorities would probably have allowed them to go further than they did. Courts
and the national Congress sometimes impeded presidential encroachments; the
story after 1983 is not exclusively about normative preferences and policy mod-
eration. While the Peronist presidents were not ideal democrats, they displayed
none of the four ex ante markers of authoritarian characteristics that Levitsky
and Ziblatt (2018) enumerate.26 They recognized the opposition as legitimate,

and they were less aggressive in pushing to expand presidential power than
leaders who have engineered executive takeovers.
The Illiberalism Index of the Varieties of Party Identity and Organization
(V-Party) project is useful for indicating the degree to which political parties
support democratic norms (Lührmann et al. 2020). The index is based on four
expert survey questions about party leaders’ use of severe personal attacks on or
demonization of political opponents; their acceptance of free and fair elections
and freedom of speech, media, assembly, and association; their acceptance of
minority rights; and their rejection of political violence. All four questions tap
essential democratic norms. The scale ranges from 0 (parties that fully embrace
liberal democratic values) to 1 (extremely illiberal parties).
Figure 7.3 shows V-Party’s Illiberalism scores for the PJ and the UCR from
1983 to 2019 and for Republican Proposal (Propuesta Republicana, PRO),
former President Mauricio Macri’s party, from 2015, when it was created, until
2019. As the conclusion to this volume (Chapter 8) shows, Argentine parties
have embraced the liberal values embodied in V-Party’s Illiberalism index to a
greater degree than the governing parties of other countries in this volume. The
PJ has adhered to liberal principles less than the UCR and PRO but more than
most governing parties represented in this volume. V-Party’s Illiberalism index
supports the narrative that follows.
The Menem administration was reasonably open toward the opposition, gen-
erally tolerant of public criticism, and somewhat accepting of congressional
checks (Cheresky 2008, 122–24; Llanos 1998; Palermo and Novaro 1996, 202–3,
256–66). Menem took some actions that, if unchecked, would have expanded his
powers, including packing the Supreme Court in 1990, successfully pushing for
a constitutional amendment that allowed him to run for reelection in 1995, and
flirting with running for a third term in 1999. However, other actors committed
to preserving democracy, including parts of the PJ, blocked him from going
down the path of eroding democracy. Moreover, Menem’s closest associates
concur that he never remotely had any intention of grabbing power; rather, he
packed the court and pushed to expand executive power as ways of dealing with
the severe economic crisis he inherited.
The PJ encouraged the sometimes violent popular mobilizations that led to de
la Rúa’s fall in December 2001 (Auyero 2007; Novaro 2006, 297). However, it did
not become an anti-democratic actor (Novaro 2002, 83). Its commitment to pre-
serving democracy helped the regime to survive in the face of a bruising social
and economic crisis. During the crisis of December 2001, the PJ accepted its role
as a loyal opposition. In December 2001, when new authorities for both cham-
bers of the national Congress were elected, Eduardo Camaño, the newly elected
president of the lower chamber, said that his task and that of Ramón Puerta, who
0.45
0.40
0.35
0.30
Illiberalism Score
0.25
0.20
0.15
0.10
0.05
0.00
83
85
87
89
91
93
95
97
99
01
03
05
07
09
11
13
15
17
19
19
19
19
19
19
19
19
19
19
20
20
20
20
20
20
20
20
20
20
PJ/FPV UCR PRO
Figure 7.3. V-Dem illiberalism scores—FPV-PJ, UCR, and PRO, 1983–2019

Note: For 2017, we use the data for Unidad Ciudadana, as all Kirchnerista candidates ran under this ticket following a split in the PJ.
Source: Lührmann et al. 2020 (variable: v2xpa_illiberal).
had been elected as provisional president of the Senate, was to guarantee the sta-
bility of the regime.27
President Duhalde generally observed democratic norms and sought to
handle the massive social outbursts of 2002 peacefully. The government gener-
ally tolerated protest movements (Novaro 2006, 295), and it always had “channels
open for negotiation with the piqueteros [the protestors]” (Godio 2006, 47).
The administration preserved a fluid communication with the opposition and
avoided radical discourse.
President Néstor Kirchner (2003–7) also exhibited restraint in how much he
attempted to undercut democratic norms, albeit with the traditional Peronist
antipathy toward mechanisms of horizontal accountability. The administration
tolerated public protest and demonstrations and avoided large-scale repression
of the piqueteros (Godio 2006, 119–23). It embraced a vigorous human rights
policy against the military dictatorship of 1976–83. In 2003, Congress removed
the last legal barriers against trials for human rights violations during the mili-
tary dictatorship.
With the election of Kirchner’s wife, Cristina Fernández, as president in 2007,
the great crisis of 2001–2 was closed. The threat of regime instability that might
have besieged Argentine democracy in the early twenty-first century receded
(Bosoer 2006; Mustapic 2005; Pérez Liñán 2007). At times, Fernández displayed
intolerance toward democratic checks and balances, but she never went far down
that path. Her party, civil society, and the partisan opposition restrained the de-
gree to which she undercut democratic checks and balances, so again, the story is
not exclusively about normative preferences.
During the 1946–55 period, the Peronist Party never did anything to prevent
the rapid descent to competitive authoritarianism. Likewise, as the democracy
of 1973–76 deteriorated severely after July 1975, the dominant faction of the PJ
did nothing to rescue the regime. In sharp contrast, party leaders blocked moves
by Presidents Carlos Menem and Cristina Fernández that would have opened
pathways toward greater executive dominance, potentially leading toward some ero-
sion of democracy. This is a tale of constraints in addition to normative preferences—
but the party imposed constraints on presidents because it no longer accepted
authoritarianism.
The Radical Party
During the first Perón period (1946–55), the Radicals became a disloyal oppo-
sition, by the late 1940s.28 They conspired against the regime, and after the 1955
coup that ousted Perón, they accepted the proscription of Peronism until 1970.
The UCR’s unwillingness to accept Peronism as a legitimate actor between the
late 1940s and 1970 helped sink the semi-democratic regimes of 1958–62 and
1963–66 (O’Donnell 1973). The UCR’s transformation to a normative prefer-
ence for democracy occurred in 1970. The signing of an agreement, “La Hora del
Pueblo” (The Hour of the People), in 1970 to work together with the Peronists to
reestablish democracy signaled a commitment to support the regime, no matter
who would win the elections. The UCR was a staunch advocate of democracy be-
tween 1973 and 1976.
From 1983 to 2001, the UCR was one of the two main parties, and it won the
1983 and 1999 presidential elections. In the post-1983 period, both parties rec-
ognized each other as legitimate and long-term actors in politics, refraining
from the use of non-democratic strategies (McGuire 1995b, 178–80). The
Radical Party has had an unwavering commitment to tolerance and democratic
pluralism, and consistently moderate policy preferences. President Alfonsín
(1983–89) was a committed democrat (Norden 1996, 80–82). He accepted con-
gressional decisions that blocked his government from undertaking reforms
that he deemed important. He inherited from the Radical tradition the idea of
a democratic pact, which played a central role in his electoral campaign in 1983
(Novaro and Palermo 2006, 530–31). This pact implied a commitment to protect
civil and political rights, strengthen democratic institutions, and limit author-
itarian powers including the influence of the armed forces (Novaro 1994, 58).
In interviews for the Oral History Archive, Alfonsín’s economic team, including
former Ministry of Economy Juan Vital Sourrouille and former secretaries and
collaborators José Luis Machinea, Carlos Bonvecchi, and Pablo Gerchunoff,
stated that consolidating democracy was their paramount goal, and that policy
decisions were affected by it.29
The other UCR president, Fernando de la Rúa (1999–2001), also was com-
mitted to observing democratic norms and constraints. He supported freedom
of speech, freedom of the press, political tolerance, and public sector transpar-
ency (Bonvecchi 2002, 125; Charosky 2002, 210–13; Novaro 2002, 85–91; see
also Freedom House 2002 and US Department of State 2002).
The UCR behaved as a loyal opposition during Peronist governments.
It showed an unyielding support for democracy when it institutionally
backed President Duhalde (2002–3) in order to guarantee the stability of
the regime (Jaunarena 2011). Moved by “a fundamental commitment to
democracy,” Radical deputies and senators supported Duhalde’s policies
(Godio 2006, 190).
This change in normative preferences also reached smaller political parties.
An example was the Record of Agreements (Acta de Coincidencias), signed on
June 7, 1984, by sixteen political parties including the UCR and the PJ (Aboy
Carlés 2004). It included a commitment to reject terrorism and any attempt to
forcibly remove the legitimately elected authorities.30
The Military
The military is the main actor outside government that can directly topple
democracy. After 1983, the Argentine military went from being an anti-
democratic actor to accepting democracy and rejecting authoritarianism.
The armed forces have moved far from the anti-democratic ideologies that
guided many officers during most of the twentieth century. Since 1990,
its leadership has adhered to the idea of civilian democratic control of the
armed forces.31 It is not clear whether the military has a normative prefer-
ence for democracy (this is possible, but we do not have enough evidence to
make the assertion), but it has moved very far from its past, when it usually
had a normative preference for dictatorship. Its acceptance of democracy has
closed the doors to coups.
Like other sectors of Argentine society, the armed forces experienced major
substantive losses during the first two decades of democracy. Many military
officials who committed human rights abuses during the dictatorship went to
jail. As of 2005, 1,070 individuals were imprisoned for human rights crimes,
the overwhelmingly majority of which were committed during the dictator-
ship.32 In addition, as president, Alfonsín drastically reduced the military
budget, from 3.47 percent of the GDP in 1983 to 1.88 percent in 1989. In the
same period, military personnel declined by 60 percent (Battaglino 2010).
The reduction in the military budget continued in the following presiden-
cies; it was slightly above 1 percent of the GDP by 2002 (World Development
Indicators, n.d.).
During the early years of the democratic regime, the military was divided
between factions that accepted civilian control and democracy, including
groups vocally in favor of democracy, such as the Center of Military Members
for Argentine Democracy (Centro de Militares para la Democracia Argentina,
CEMIDA),33 and minority sectors that remained wedded to the authoritarian
tradition and engaged in uprisings in 1987 and 1988 (Jaunarena 2011; Norden
1996). Another army revolt took place in late 1990, but loyal army units re-
pressed the insurgents. After Menem quelled the last major military rebellion in
1990, the armed forces became a less relevant political actor (Norden 1996). By
the early 1990s, the military was under civilian control and began to acknowl-
edge excesses during the “dirty war.” Military extremism was buried. During the
protests of December 2001, the military refused to suppress the protests without
an explicit law from the National Congress. Godio (2002, 126) states that by then
the armed forces were unified in the preservation of democracy. The military is
more constrained now by other democratic actors than it was between 1930 and
1983, but these constraints imposed by other actors are themselves a product of
changing attitudes about democracy and authoritarianism.
Organized Labor
In Argentina, organized labor was frequently an important actor in the fate of

political regimes. Before 1983, most of the organized labor movement was at best
indifferent toward democracy. Labor mobilized against the semi-democratic
regimes of 1958–62 and 1963–66 and supported coups against them. Most of
the movement was wedded to Perón (1946–55, 1973–4) despite his turn toward
a repressive competitive authoritarian regime in the 1940s and his ambiguous
practices toward democracy in 1973–4. Incessant demands for higher wages,
coupled with minority radical left-wing mobilizations that included countless
factory occupations, helped galvanize the opposition to the democratic regime of
1973–6 (Torre 1983). The Peronist right wing of the labor movement supported
the death squads, enabling a severe degradation of democracy as human rights
abuses escalated.
Workers suffered substantial economic losses during the Alfonsín period.
The average income of the employed active population in Greater Buenos Aires
declined 24 percent between 1986 and 1990 (ECLAC 1993). Workers’ economic
losses were deeper than those they experienced during the military dictatorship.
Contrary to the argument that democracy favors income redistribution to the
poor (Acemoglu and Robinson 2006; Boix 2003), exactly the opposite happened
in Argentina during the first two decades of democracy: income inequality
increased sharply and the median income declined.
Despite these losses, organized labor was deeply committed to preserving
democracy. One month after Alfonsín’s inauguration, in January 1984 the
unified CGT issued a document that captured labor’s new position on repre-
sentative democracy. “The rule of law, liberty guaranteed by the legal system,
responsible pluralism, and institutionalized participation constitute the only
path by which economic development is viable, social justice is possible,
and the realization of the human person is guaranteed. For this reason, the
labor movement must be a zealous guardian of democratic stability for all
Argentines” (quoted in García Lerena 2007, 240–41). In March 1984, the CGT
called upon business associations, parties, and youth and religious organi-
zations to integrate a broad national front to help protect democracy (Senén
González and Bosoer 2009, 235). McGuire (1997, 21) writes that by the end of
the Alfonsín period, the main contending factions within the labor movement
“acquired an instrumental stake in the survival of elections and legislative ac-
tivity” and that the leader of the CGT, Saúl Ubaldini, was committed to democ-
racy “was beyond any serious dispute.” In the words of Ramón Baldassini, one
of the four Secretary Generals of the CGT, Argentines needed to become con-
scious of the fact that an institutional break would hurt everyone in the long
run and workers in particular.34
Labor vigorously mobilized against the Alfonsín government. Even during

strikes and demonstrations against the government, however, organized labor
stressed its commitment to democracy. In a gathering in the Plaza de Mayo in
May 1985, Ubaldini stated that there would not be a coup because workers would
stop anyone who might try (quoted in García Lerena 2007, 250). In November
1984, José Luis Castillo, the secretary general of the Union of Ship Technicians
of the Argentine Republic (Sindicato de Conductores Navales de la República
Argentina, SICONARA) expressed concern with rumors about a coup and stated
that it was necessary to defend democracy, no matter who the president was.35
Some of labor’s criticisms of the Alfonsín government were based on the idea
that democracy was not deep enough. These criticisms called for deepening de-
mocracy rather than rejecting it to prioritize social or economic outcomes or
corporatist interests. A constant in these declarations was the claim that workers
were crucial in the fight for getting democracy back and that they were now cru-
cial in defending it.36
Labor commitment to democracy was also evident in cross-sectional actions
such as the “Democratic Compromise in the Face of the Situation of Rebellion
and Threat against the Constitutional Order” (Compromiso democrático ante la
situación de rebeldía y amenaza del orden constitucional) signed in 1987 by the
CGT, political parties, the Argentine Industrial Union, a powerful business as-
sociation (Unión Industrial Argentina, UIA), and the government in response to
the military uprisings (more on this below). The signatories endorsed democ-
racy as the only option for Argentines (García Lerena 2007, 264).
Labor’s defense of democracy was especially important during the military
uprisings. During the Easter 1987 uprising, the CGT declared a strike for an
unlimited time starting the following Monday if the rebels did not surrender
(Pucciarelli 2006, 136). An official CGT resolution stated that the Argentine
people had to be united against any threat to democracy and defend it without
hesitation.37 As labor leader José Pedraza stated in an interview, CGT leaders
“monolithically defended democracy . . . regardless of our differences with
Alfonsín, with the Radical Party, because we stood by democracy.”38 As Luis
Sartori highlighted in a newspaper article published right after the events, that
was the first time that a non-Peronist government received unanimous support
from the unions against a military rebellion.39 This unanimous support was re-
peated during the next military uprising in January 1988.40
In June 1988, a group of union leaders led by Ubaldini, who opposed the Due
Obedience law that gave amnesty to human rights violators who had not yet
been prosecuted, issued a joint statement saying, “The armed forces . . . [should]
subordinate themselves to the democratic system and guarantee the people that
nothing will stand in the way of the prosecution of those who . . . violated human
rights” (quoted in McGuire 1992, 54–55). Writing at a time of severe economic
problems that had terrible consequences for many workers, Ranis (1992, 6) stated
that “Argentine workers’ commitment to and support for democracy seemed to
reach a new plane in the light of the contrast between the military Proceso and
the new democracy.”
During the economic and financial meltdown of 2001–2, although labor or-
ganizations protested and criticized the government, they did not oppose the
democratic regime. In a massive demonstration in downtown Buenos Aires, co-
organized by unions and labor organizations in August 2001, Hugo Moyano, the
leader of the most combative faction of the CGT, urged Argentines to use the “le-
thal weapon” of elections to express their dissatisfaction with the government.41
Popular Mobilizations
During the military rebellions, massive demonstrations in support of democ-

racy showed citizens’ commitment to the regime (Jaunarena 2011). The first big
pro-democracy, anti-coup demonstration took place in Córdoba province on
Holy Thursday, April 16, 1987, one day after the first signs of military revolt be-
came evident in Córdoba. Local parties, together with social and professional
organizations including the student movement and the local branch of the CGT,
organized a popular demonstration in defense of democracy and a multiparty
condemnation of the revolts. Famous singers, actors, and actresses took part in
many of the popular demonstrations, called by their professional associations
(Pucciarelli 2006, 123–24). Alfonsín later said, reflecting on the Easter uprising,
that his only force for deterrence in this dramatic situation was the gigantic and
patient crowd covering Plaza de Mayo and other public squares (quoted in Masi
2014, 162). As Gabetta (quoted in Pucciarelli 2006, 134) stated in 1987, the crowd
was willing to risk their lives for the democratic regime, showing that Argentines’
attitude toward democracy had changed.
The economic crisis of 2001–2 generated widespread social protest. However,
these protests were different in kind and political impact from the radicalized
labor and student mobilizations of the 1973–76 period. In the 1973–76 period,
radicalized labor groups and student organizations occupied countless factories
and universities, adding to the tumult and the sense of chaos that weakened the
democratic regime. These movements were at best indifferent to the fate of de-
mocracy; some favored the installation of a revolutionary socialist regime, and
others engaged in behavior that destabilized democracy.
After 1983, even most radical movements protected democracy. An ex-
ample of this was the emergence of organizations of unemployed that became
known as “organizaciones piqueteras” (picketer organizations), due to their
utilization of roadblocks (Svampa and Pereyra 2003, 25). These movements of
unemployed and informal poor workers emerged in the late 1990s in the face
of very high unemployment and a weak public safety net. These mobilizations
initiated a dynamic of protest for social benefits and provisions, including the
use of roadblocks as their main tool. The protests intensified as social conditions
deteriorated and the government failed to provide adequate responses to the
demands (Garay 2016, 169–71).
Although the piqueteros adopted a repertoire of contentious action,
their policy demands seldom acquired overtones against democracy. The
piqueteros employed confrontational tactics by blocking roads and highways,
vandalizing toll booths, blocking the pay booths for subways, and encour-
aging people to ride the subways without paying (Cheresky 2008, 189). In
2004 some groups committed violent acts, including an attack on a police
station (Botana 2006, 140). However, most factions were not anti-system
(Godio 2006, 307, 319).
The protest movements in 2001–2 never questioned democracy as a political
regime. Their main demands were oriented to the state (Schuster et al. 2006). The
movements expressed discontent with the political establishment but not against
democracy. On the contrary, they demanded more democracy in the form of
better representation or the possibility of intervening more directly in public af-
fairs. Some neighborhood assemblies referred to elections as an opportunity for
people to use their vote to kick everyone out (Dinerstein 2004, 262).
Many social movements involved in the protests aimed at recovering the
participatory component of democracy, faced with a democracy that they con-
sidered insufficiently democratic (Natalucci 2011, 198). The protests were thus
aimed at defending and not attacking democracy. In the words of one partici-
pant, after De la Rúa’s call for the “state of siege,” “I lived through the junta (mili-
tary government); I know what it means. . . . [P]‌eople saw this as the first step in
the return to political repressions . . . A reversal of rights that we were not willing
to lose” (in Onuch 2014, 105).
Business Associations
Business associations openly supported the 1930, 1955, 1966, and 1976 coups
(Sidicaro 2002). Their behavior changed to pro-democratic after 1983, partially
in response to the failed economic policies of the dictatorship.
Many business sectors experienced heavy setbacks during the periods from
1983 to 1991 and from 1998 to 2003. According to Central Bank data, the main
index for the Argentine stock market, Merval, fell more than 60 percent between
February 2000 and October 2001.42 If we consider the devaluation of the peso,
the drop in real terms during that period was 90 percent (to one-tenth of its
value).43 This collapse of the Argentine stock market negatively affected many
business and elite interests.
Despite these economic losses for business, the experience of the last mili-
tary dictatorship brought about a change in business’s preferences for democ-
racy. Acuña (1995, 232) identifies the democratic transition as a turning point
with respect to previous patterns of political struggle for business interests. He
observed a change in the meaning that authoritarianism and democracy have
for their economic interests. Business interests ceased seeing dictatorships as
friendly and predictable. Dictatorships might pursue ill-conceived policies with
negative impacts for the main economic actors or perilous “adventures” such as
the Falkland/Malvinas War (Acuña 1995, 268–69). Because of the failures of the
economic policies of the dictatorship of 1976–83, business stopped perceiving
the military as an agent of modernization. This led to support for democracy
in the months leading to the elections of 1983 (Pucciarelli 2006, 216). The busi-
ness sector’s bet on the democratic regime seemed to be a long-term one (Acuña
1995, 272). In 1985, the “Group of 11,” made up of the CGT and the main busi-
ness associations, stated in a public communication that democracy is a patri-
mony of the whole nation and an unquestionable value.44 Business associations
defended the Alfonsín government during all of the military uprisings (Acuña
1995, 269; Beltrán 2006, 217). Business leaders viewed the leaders of the mili-
tary rebellions with suspicion and skepticism. Their modes of action, populist
slogans, and references against “imperialism” made them a risky and uncer-
tain alternative for business. Associations such as the Argentine Rural Society
(SRA), the Argentine Industrial Union (UIA), and the Argentine Chamber of
Commerce (Cámara Argentina de Comercio, CAC) spoke out in favor of democ-
racy during the military crises (Acuña 2004, 180–81). Acuña (1995) argues that
business changed its normative preferences regarding the political regime, and
not only its instrumental policy preferences after 1983 in response to the failures,
closed nature, and erratic character of the 1976–83 regime.
An act that illustrated the support for democracy by opposition parties,
unions, and business organizations was the signing of the “Democratic
Compromise in the Face of the Situation of Rebellion and Threat against the
Constitutional Order” in response to the military uprising of April 1987. The
document expressed the determined resolution of the signatories to defend the
Constitution and democracy as the “single life destiny of the Argentine people.”45
The document was signed by representatives of most political parties including
the Radical Party, the Peronist Party, the Intransigent Party, the Union of the
Democratic Center (Unión del Centro Democrático, UCEDE), the Christian
Democratic Party, and many provincial parties. Representatives of the CGT and
many business organizations, including the SRA, the UIA, the Construction
Union, the CAC, and the Argentine Rural Confederations, also signed.46
By 2001, thus, all the main actors had accepted democracy. No actors have
an explicit preference for dictatorship, and none has attempted to form an au-
thoritarian coalition since the failed coups toward the end of the Alfonsín
government.
The Decline of Extremist Actors and the Rise

of the Human Rights Movement
The changes in attitudes toward democracy and authoritarianism were accom-

panied by a concomitant demise of radicalism in Argentine politics. Powerful
actors with extremist agendas no longer exist. The revolutionary left was annihi-
lated during the “dirty war”. Most individuals who came from the revolutionary
left and survived the dictatorship underwent political and personal conversions.
Many became staunch supporters of democracy (Ollier 2009). The demise of
the revolutionary left and of radical sympathizers such as the Peronist Youth
(Juventud Peronista) and far left-wing unionism lowered the stakes of demo-
cratic politics. Conservative and centrist actors no longer feared politically moti-
vated assassinations and kidnappings, factory takeovers, and the sense of chaos
and disorder that permeated the 1973–76 period.
The extremist right wing, which included death squads that killed hundreds
of people before the March 1976 coup, also weakened greatly, discredited by
its association with a disgraced regime. The radical right-wing temptation—
the idea that Argentina could recover its golden era through right- wing
authoritarianism—died. The defeat in the 1982 Falkland/Malvinas War, the mil-
itary regime’s mishandling of economic policy, the international notoriety and
condemnation that surrounded the state terror of the military regime, and the
domestic repudiation of the human rights violations committed by the armed
forces during the “dirty war” discredited the far right. After 1983, other political
actors rejected extremist policy positions and intransigent behavior.
The end of extremism, which was closely associated with changing attitudes
toward democracy, had profound consequences for democratic survival. In
periods of economic downturn, some of which have been severe in the post-1983
period, most people, companies, and sectors have suffered losses, but democracy
has given them an institutionalized, peaceful way of fighting for their interests
without resorting to coups or violence.
During the dictatorship, some courageous individuals created new human
rights organizations. They established one of the most innovative, vociferous,
and effective human rights movements in the world. They were a key part of the
opposition to the dictatorship, and a consistent advocate thereafter clamoring for
transitional justice. Although thirty-eight years have now lapsed since the end
of the last dictatorship, these organizations have continued their work, thereby
continuously activating a message on behalf of human rights and, by extension,
on behalf of democracy (Sikkink 2008).
Understanding Changes in Actors’ Preferences
The deep failures of the military dictatorship of 1976–83 were a major cause of
the reorientation of actors’ preferences about the political regime. Repressive
failed dictatorships do not always generate change in actors’ preferences toward
democracy (Roberts 1998), but they sometimes do (Bermeo 1992). In this re-
spect, the reorientation of many Argentine actors toward democracy after the
1976 coup followed a path similar to what Bermeo describes in her chapter on
Timor-Leste. A process of political learning occurred. The building of collective
memory through human rights trials resulting in the imprisonment of hundreds
of abusers, human rights museums and other public spaces, a vigorous human
rights movement, and a vast outpouring of arts, journalism, and academic work
on the abuses of the dictatorship has constantly renewed a message about human
rights and democracy.
An extensive literature has documented that international influences, espe-
cially regional influences, affect political regimes. Two international influences
supported the demise of extremist actors and a stronger commitment to democ-
racy and greater political moderation among most of the rest. First, US policy to-
ward Latin America changed after 1977. For the most part, from 1977 until 2017,
the US supported democracy in Latin America, although the periods from 1981
to 1985 and from 2017 to 2021 were exceptions. Except for these two periods,
the US generally withdrew support for right-wing authoritarians. Because actors
knew that the US would oppose coups against democracy, the cost of attempting
coups increased. A decisive turning point in the US’s relationship with Argentina’s
right-wing authoritarians came when President Carter signaled clear oppo-
sition to the military dictatorship of 1976–83. Another came when President
Reagan sided with the United Kingdom when the Argentine dictators went to
war against the US’s close ally in 1982. US opposition to the generals’ bellicose
miscalculation reduced the space for Argentina’s right-wing authoritarians. US
support for Argentine democrats bolstered the pro-democracy camp.
Second, the transformation of the Argentine left was influenced by a sim-
ilar development on the left in Brazil, Chile, and Uruguay, as well as in much
of Europe. The left began to reject revolutionary socialism and to embrace lib-
eral democracy as a necessary part of a desirable political order (Ollier 2009). In
Uruguay, Chile, and Brazil, the devastating experiences of military dictatorships
pushed the left toward democracy. The left in these countries and the democratic
left in Europe influenced leftist Argentine intellectuals, politicians, artists, and

political activists. The crushing of dissent in Poland in 1981 laid bare the bank-
ruptcy of Soviet-style communism. The motto of the Argentine human rights
movement, “nunca más,” meaning, “never again [will we allow such depraved
human rights abuses],” could also apply to most of the left’s attitude about revo-
lutionary socialism: never again would these individuals and organizations em-
brace revolutionary socialism.
A Few Alternative Explanations
Some scholars argue that institutional constraints, rather than actors’ norma-
tive preferences and their location on the moderate–radical policy scale, help
explain the survival of democracy. Institutional constraints might be sufficient
to protect democracy in countries with long-established democracies and strong
institutions. However, Argentina has relatively weak democratic institutions
(Levitsky and Murillo 2005; Spiller and Tommasi 2007). The courts (Helmke
2005; Larkins 1998) and Congress have sometimes imposed limits on presi-
dents, and in so doing, they have helped defend democracy against presidential
encroachments. But by themselves, they do not provide a convincing expla-
nation for the stabilization of Argentine democracy. They were not significant
actors in defeating the military rebellions of the 1980s. Congress sometimes
limited initiatives by the UCR and Macri administrations, but for the most part
it was highly deferential to the PJ presidents. As Jones and Hwang (2005, 127–
28) summarized, “The majority party leadership in the Chamber and Senate is
in most instances a faithful servant of the president.” Moreover, as Jones et al.
(2002) and Jones and Micozzi (2013, 72) noted, the combination of very short
legislative careers (usually, one term only for federal deputies), governors’ con-
trol over deputies, and ample party switching after elections (usually in the direc-
tion of the president’s coalition) weakens legislative controls over the executive.
Notwithstanding occasional exceptions, the courts have been generally defer-
ential to presidents (Berbecel, 2022; Helmke 2005; Spiller and Tommasi 2007).
The hypertrophy of presidential power during the Menem period led O’Donnell
(1994) to coin his famous concept, “delegative democracy,” one of whose central
features was the weakness of mechanisms of horizontal accountability.
In the post-1983 period, some institutional constraints have been stronger
than they were in the democratic experiences of 1946–51 and 1973–76. The fact
that the governing party has not always had a majority in Congress has enhanced
congressional checks and balances. The fact that no party consistently wins the
presidency has been salutary for democracy. Nevertheless, the ability of many
provincial governors to construct competitive authoritarian provincial enclaves
(Gervasoni 2018) and the weakness of Argentina’s court system in checking

presidents indicate that institutions alone are unlikely to suffice to preserve de-
mocracy. The ability of presidents and prime ministers in seemingly once solid
democracies such as Venezuela until the 1990s, Hungary until 2010, and Poland
until 2015 to overrun democracy points in the same direction.
Some excellent scholars included in this volume (and others) posit that a bal-
ance of power constrains powerful actors to abide by democratic rules of the
game and that this balance of power explains democratic survival. We agree that
a balance of power is generally good for democracy. Dominance by one party or
individual makes executive takeovers easier, and it could make military coups
more likely if powerful actors strongly oppose the government. However, after
the fall of the De la Rúa government in 2001, except for the 2015–19 interlude,
the Peronists have dominated the electoral and political landscape against a
highly fragmented opposition. In the context of generally deferential courts and
a national congress that is not strong, it is difficult to argue that the opposition
was powerful enough to have kept a president who really wanted to undermine
democracy from doing so. The fact that courts and Congress have occasionally
limited presidential encroachments does not show that they have consistently
been decisive constraints in safeguarding democracy. In contexts of fairly weak
institutional constraints and clear electoral dominance, power-hungry presi-
dents who are bent on eroding democracy are difficult to stop.
Could public opinion favorable to democracy explain its survival? In the Latin
American Public Opinion surveys, Argentines have consistently expressed more
favorable attitudes about democracy than respondents in most Latin American
countries. For example, in 2018–19, 71.1 percent of Argentine respondents
agreed that “Democracy may have problems, but it is better than any other form
of government.” In Latin America, this was the third highest percentage, be-
hind only Uruguay (76.2 percent) and Costa Rica (72.4 percent) (Zechmeister
and Lupu 2019, 12). Moreover, at critical times such as the military rebellions
discussed above, citizen mobilization on behalf of democracy helped a great
deal. Certainly, favorable public opinion has been a substantial asset for democ-
racy. But it is hard to know how much weight to give it. Venezuelans expressed a
high attachment to democracy as their country became more and more author-
itarian, so citizen support in the usual survey questions certainly does not guar-
antee the survival of democracy. More generally, it is not easy for voters to force
governments to remain faithful to democracy (Graham and Svolik 2020).
Gibson (1996) and Ziblatt (2017) argued that strong conservative parties help
stabilize democracy by protecting the interests of conservative elites such as
business groups. This argument is generally right, but it does not help explain the
survival of democracy in Argentina. From 1983 to 2015, Argentina had nothing
that resembled a strong conservative party.47 The most successful conservative
parties in presidential elections were the UCEDE in 1983 with 0.4 percent of the
vote and again in 1989 with 6.5 percent, MODIN in 1995 with 1.8 percent, Acción
por la República in 1999 with 10.2 percent, and Recrear in 2003 with 16.4 percent
and in 2007 with 1.4 percent. No conservative party had a presidential candidate
in 2011. Democracy can endure despite the lack of a powerful conservative party.
Unless the major parties are committed to democracy, it is doubtful that it can
survive powerful extremist actors committed to the destruction of democracy, as
Argentina experienced from 1973 to 1976.
The absence of a solid conservative party was not a hindrance to democracy
despite the fact that the main conservative actors suffered large losses. The mili-
tary, business, and the Church hierarchy (which was mostly a conservative actor
in Argentina from 1976 to 1983)48 accepted large losses under democracy. As
noted previously, many military leaders ended up in jail, military budgets were
slashed, and the military was downsized. Most business interests faced losses
during the economic crises of 1983–90, 1998–2002, and 2018–20. Agricultural
exporters were subjected to large export taxes under the government of Cristina
Fernández de Kirchner (2007–15). Most of the leadership of the Argentine
Catholic Church supported the military coup of 1976 and the ensuring dicta-
torship. Under democracy, the Church lost important political battles in 2006,
when the Law of Integral Sexual Education passed; in 2010, when Argentina be-
came the first Latin American country to pass a bill legalizing gay marriage at
the national level; in 2012, when it passed one of the most progressive laws in
the world in support of the rights of transgender individuals; and in 2015, when
a new civil code affirmed the equal rights of all individuals regardless of gender
and sexual orientation. Yet the Church has accepted democracy and generally
embraced it since 1983.
A modernization explanation can explain the survival of democracy after
1983, but it cannot explain why democracy has fared better since 1983 than it
did between 1973 and 1976, and it offers at best a weak explanation for why de-
mocracy has fared better since 1983 than in the breakdowns of 1930, 1951, 1962,
and 1966. Although Argentina’s level of development has long been favorable
to democracy, the initial level per capita GDP in 1983 was slightly lower than
it was in 1973. Per capita GDP was $6,717 in 1983, 3 percent lower than the per
capita GDP of $6,947 at the beginning of the previous democratic regime in 1973
(World Development Indicators, n.d.).
Class theories of democratization based on the power of the working class do
not explain Argentina’s democratic stability since 1983. The organized working
class is smaller and less powerful now than it was during Argentina’s competi-
tive regimes of 1946–51, 1958–62, 1963–66, and 1973–76. The share of the eco-
nomically active population engaged in manufacturing, mining, transportation,
and construction declined from 28 percent in 1970 to 22 percent by 2005 (ILO
1975; LABORSTA, n.d.). Union density declined steeply from 42.5 percent in
1954 (Doyon 1988, table 5) to 24.4 percent for 1990–95 (ILO 2002). Therefore,
Argentina’s post-1983 democratic stabilization is not a result of a larger, more
powerful working class.
Theories of democratization based on income inequalities also fail to explain
Argentina’s transformation from repeated breakdowns to stable democracy after
1983. Argentina’s income distribution was relatively equal between 1960 (and
almost certainly well before then) and 1976. From 1960 until 1976—the first
decades for which hard data are available—Argentina’s Gini coefficient for in-
come distribution was almost always below 0.40 and was as low as 0.33 (1972),
one of the lowest figures ever recorded in Latin America. Income inequalities
became much worse in Argentina in the decades after the 1976 coup. From 1976
until 2003, Argentina experienced the greatest increase in inequality of any
country in Latin America. The Gini index increased from 34.7 in 1973 (the inau-
gural year of the last democratic regime before the current one) to 52.8 in 2003.
Few countries in the world outside the former communist countries experienced
such a staggering increase in inequality in this time. When the main actors were
committed to democracy and there were no extremist actors, the regime could
withstand a profound increase in inequality (SEDLAC 2018).
Conclusion
In Latin America, democracy has infrequently broken down since the beginning
of the third wave in 1978. Democracy has survived in most countries despite for-
midable obstacles and deep economic crises. The specific story behind the sur-
vival of democracy varies from country to country, but a common story is that
more actors are committed to democracy than before, fewer authoritarian actors
abound, extremist actors have been less powerful, and US policy has been largely
favorable to supporting democracy since Jimmy Carter’s presidency (1977–81)
(Schenoni and Mainwaring 2019).
In Argentina, democracy survived despite a past history of coups and strong
authoritarian actors, and despite three severe economic crises and several mil-
itary uprisings. Because substantive outcomes were terrible for many actors
during the first two decades of democracy when it was more likely to be vul-
nerable, it is impossible to explain the survival of democracy by claiming that
actors benefited in substantive terms. And because factions of the military were
probably willing to launch a coup, it is clear that actors did not merely accept de-
mocracy because they had no choice. Some of the actors and millions of people
who experienced the deepest losses in the 1983–91 period mobilized in defense
of democracy when the military rebellions occurred.
Democracy survived in part because of the demise of extremist actors. With

less at stake, there was less reason to destroy democracy. In addition, a funda-
mental transformation occurred in several key actors’ beliefs regarding de-
mocracy after the last military dictatorship. Some actors came to believe that
democracy was worth defending even if its substantive results were dismal.
Other actors were less committed to democracy, but they were more committed
than they had been in the past to not flagrantly violate democratic norms even
when they had the power to do so. This normative preference and the demise of
radical extremist actors has helped preserve democracy.
Notes
* A few paragraphs of this chapter come from Mainwaring and Pérez-Liñán 2013.
We are grateful to Fernando Bizzarro, Michael Coppedge, Candelaria Garay, Steve
Levitsky, Raúl Madrid, Tarek Masoud, Aníbal Pérez-Liñán, Luis Schiumerini, Maggie
Shum, Jazmín Sierra, Guillermo Trejo, Deborah Yashar, and colleagues at the confer-
ence on Democracy in Hard Places that took place in May 2019 at the Ash Center for
Democratic Governance and Innovation of the Harvard Kennedy School for helpful
comments. Thanks to Vicky De Negri and Luis Elizondo for research assistance.
1. A military coup ended the presidency of Juan Perón in 1955. In our view, democracy
had broken down by 1951 because of severe infringements of opposition rights and a
grossly uneven playing field. Thus, we situate the breakdown as occurring four years
before the coup that deposed Perón.
2. The exposition that follows focuses on coups, but the logic applies equally to execu-
tive takeovers.
3. Measured in annual percentage (indicator: NY.GDP.MKTP.KD.ZG).
4. For a similar logic with voters, see Graham and Svolik (2020).
5. Frey et al. (2004, 381) define procedural utility as “the well-being people gain from living
and acting under institutionalized processes as they contribute to a positive sense of self.”
6. “Argentine Falklands War Troops ‘Tortured by Their Own Side,’ ” BBC News,
September 14, 2015, https://www.bbc.com/news/world-latin-america-34252025.
7. Both movies won Oscar awards for Best Foreign Language Film as well as many other
international film awards.
8. For example, a best-selling book by renowned journalist Horacio Verbitsky (1995) re-
ported confessions of naval officer Adolfo Scilingo that the Argentine military had dumped
the live bodies of countless political prisoners into the Atlantic Ocean on “death flights.”
9. “7 sitios para recordar y volver a decir: ‘Nunca más’ [7 sites to remember and say once
more: ‘Never again’].” Ministerio de Cultura, March 23, 2019, https://www.cultura.
gob.ar/espacios-para-no-olvidar-7-sitios-de-horror-y-resistencia_7297/.
10. For evidence that human rights museums make individuals, in the short term, more
supportive of democracy and transitional justice, and less supportive of authoritari-
anism, see Balcells et al., 2022.
11. Measured in annual percentage (indicator: NY.GDP.MKTP.KD.ZG).

12. Measured in constant 2010 US dollars (indicator: NY.GDP.PCAP.KD).
13. Annual growth in consumer prices.
14. GDP per capita growth (annual %) (NY.GDP.PCAP.KD.ZG).
15. Authors’ translation. All translations in this chapter are ours.
16. Thanks to Luis Schiumerini for pointing out this survey result.
17. The formula is (14-FH score)/12, where FH score is the combined score for civil lib-
erties and political rights. Although the scale is the same, the scores are not exactly
commensurable; Freedom House has far more scores equal to or approaching 1 than
V-Dem. V-Dem counts the first year of democracy as 1984, thus accounting for the
sharp discrepancy in the 1983 scores.
18. Measured in constant 2010 US dollars (indicator: NY.GDP.PCAP.KD).
19. The government of Patricio Aylwin in Chile (1990–94) and the Chilean courts vigor-
ously punished a large number of human rights abusers. However, their efforts were
not as publicized as those in Argentina, and they never induced a military rebellion
that could have resulted in a coup. In most other Latin American countries, early
efforts at transitional justice were timid (González Ocantos 2016).
20. Archivo de Historia Oral de la Argentina Contemporánea, Instituto de Investigaciones
Gino Germani, http://iigg.sociales.uba.ar/cdi-2/archivos/.
21. Archivo del Sindicalismo Argentino “Santiago Senén González” (ASASG),
Universidad Torcuato Di Tella, https://www.utdt.edu/ver_contenido.php?id_conten
ido=2280&id_item_menu=4559.
22. This includes all the public declarations and speeches available at the library of the
CGT for the period 1981–90 (although there are only a couple of documents for the
last three years).
23. “Iniciativa de Labaké para recrear la estructura del movimiento peronista [Labaké’s
initiative to recreate the structure of the Peronist movement]” Tiempo Argentino,
March 3, 1985.
24. Abal Medina served as chief of cabinet for President Cristina Fernández de Kirchner
(2011–13) and national senator (2011–17). His father was one of the founding leaders
of the Montoneros.
25. Along similar lines, Tcach and Quiroga (2006, 13) wrote that “[t]‌he dictatorship
taught the society . . . to value the preservation of democracy. . . . Argentine society
learned the lesson well: the rejection of political violence that undermines and annuls
institutional legitimacy and the rule of law.”
26. Levitsky and Ziblatt (2018, 23–24) propose four key indicators of authoritarian
orientations: (1) “Rejection of . . . democratic rules of the game,” (2) “Denial of the
legitimacy of political opponents,” (3) “Toleration or encouragement of violence,” and
(4) “Readiness to curtail civil liberties of opponents, including media.”
27. Quoted in “El nuevo Parlamento: los límites del justicialismo. La fortaleza
institucional no disimula debilidades internas [The new Parliament: the limits of
Justicialism. Institutional strength does not conceal internal weaknesses].” Clarín,
December 6, 2001, https://www.clarin.com/politica/fortaleza-institucional-disim
ula-debilidades-internas_0_rkVxzPUgRFl.html.
28. Linz’s (1978b) concept of “disloyal opposition” referred to actors that wanted to un-
dermine a democratic regime. Because Perón installed a repressive competitive au-
thoritarian regime, we use the term “disloyal opposition” here to refer to the UCR’s
attitude toward this regime, not toward democracy per se.
29. Oral History Archive interview recorded during July 2005 (first part, first cassette).
30. Martín Prieto, “Peronistas y radicales firman en Argentina un remedo de los ‘pactos
de la Moncloa [Peronists and Radicals sign in Argentina an imitation of the Moncloa
pacts].” El País, August 6, 1984, https://elpais.com/diario/1984/06/08/internacional/
455493620_850215.html.
31. A few prominent military leaders published memoirs or gave interviews that
endorsed the intrinsic value of democracy. See Ballester (1996, 14) and the interview
with former president Alejandro Lanusse (1971–73; Lanusse 1988, 59–63).
32. Procuraduría de Crímenes contra la Humanidad, n.d. [Attorney’s Office for Crimes
against Humanity] “Estadísticas” (Statistics), https://www.fiscales.gob.ar/lesa-
humanidad/?tipo-entrada=estad%C3%ADsticas (accessed April 2021).
33. See “Nucleamiento de militares retirados [Grouping of retired military officials].”
Clarín, November 18, 1984.
34. “Ramón Baldassini se define como un socialdemócrata [Ramón Baldassini defines
himself as a social-democrat]. La Razón, December 26, 1984.
35. “Preocupa a la C.G.T. la desestabilización [CGT worried by destabilization].”
Clarín, November 6, 1984. Castillo later served as a federal deputy from the Peronist
(Justicialist) Party for the province of Buenos Aires from 1987 to 1999.
36. “ ‘Tenemos una democracia sólo formal [We have only a formal democracy].” Tiempo
Argentino, April 27, 1984; “Casella dialogó con el titular de la CGT [Castella con-
versed with the head of the CGT].” Clarín, April 27, 1984; “Los trabajadores aspiramos
a una democracia real, con justicia social e independencia económica [“We workers
aspire to a real democracy, with social justice and economic independence”].
Tiempo Argentino, April 30, 1984; “CGT: Multitudinaria Concentración [CGT: Mass
Concentration].” May 2, 1984; “CGT: Rechaza la advertencia oficial [CGT: Rejects the
governmental warning].” Clarín, November 22, 1984.
37. “Resolución del Comité Central Confederal [Central Confederate Committee’s reso-
lution].” CGT, March 25, 1987.
38. Third session, minutes 56:50–58:00. Interviewed by members of the Oral History
Group at IIGG Archivo de Historia Oral de la Argentina Contemporánea.
39. Luis Sartori, “Un respaldo sin fisuras [Unfailing support].” Clarín, April 20, 1987.
40. Ricardo Roa, “Acordes y disonancias frente a la crisis [Harmonies and dissonances in
the face of the crisis].” Clarín, December 12, 1988.
41. Quotes published in “Las CGT pidieron derrotar en las urnas el modelo económico
[The CGTs call for the defeat of the economic models in the ballot box].” La Nación,
August 30, 2001, https://www.lanacion.com.ar/331412-las-cgt-pidieron-derrotar-
en-las-urnas-el-modelo-economico.
42. Banco Central de la República Argentina [Central Bank of the Argentine Republic],
Informe Monetario Mensual [Monthly monetary report] (March 2000 and November
2001; accessed August 2, 2021), http://www.bcra.gob.ar/PublicacionesEstadisticas/

Informe_monetario_mensual.asp.
43. Marc Garrigasait, “Qué hizo la Bolsa en Argentina durante el corralito y qué nos
enseña [What the Argentine Stock Market did during the ‘corralito’ and what that
teaches us].” El blog de Marc Garrigasait (blog), August 6, 2012, https://investorsco
nundrum.com/2012/08/06/que-hizo-la-bolsa-en-argentina-durante-el-corralito-y-
que-nos-ensena/.
44. “Apoyo a la democracia de gremios y empresarios [Support for democracy from
unions and businessmen].” Clarín, April 20, 1985.
45. Quoted in Martín Prieto, “Alfonsín logra personalmente la rendición de los
sublevados en Campo de Mayo [Alfonsín personally achieves the surrender of the
rebels of Campo de Mayo].” El País, April 20, 1987, https://elpais.com/diario/1987/
04/20/internacional/545868006_850215.html.
46. National broadcast of the signature ceremony by Radio Nacional. Audio available
at the La Plata University archive, April 24, 1987, http://sedici.unlp.edu.ar/handle/
10915/44452.
47. President Carlos Menem (1989– 99), a Peronist, governed on the center- right.
However, the PJ was not a conservative party, and it returned to its left-of-center roots
from 2002 on.
48. See Mignone (1986).
8
Why Democracies Survive
in Hard Places
Scott Mainwaring
This volume addresses a puzzle that has important theoretical consequences for
how social scientists and historians understand democracy: why does democ-
racy sometimes survive in very inauspicious conditions?* Because many struc-
tural and cultural conditions are unfavorable in these contexts, in this chapter
I focus on a debate about institutions and actors.1 Do democracies survive be-
cause narrowly self-interested actors are constrained by a balance of power and
institutions, or do they also survive in part because some actors are committed
normatively to democracy? We believe that this is the first volume to take this
puzzle up directly and in great detail.
The volume does not offer definitive answers, but it brings the debate about
these questions to the fore in new ways. No analysis of why democracy survives
can ignore actors’ instrumental interests and behavior, formal institutions, or the
balance of power, but I argue that actors’ normative preferences about democracy
and dictatorship also help explain the longevity of democracy in hard places.2
Actors have very different attitudes and preferences about democracy and author-
itarianism, and these preferences about the regime are important for sustaining
democracy or leading to its demise. Actors that are normatively committed to de-
mocracy can help sustain it even in difficult circumstances. Conversely, illiberal
rulers and other actors can often tear it down in less institutionalized regimes.
In Argentina, India, South Africa, and Timor-Leste, a repudiation of the
terrible past generated a new appreciation of democracy that has helped these
regimes survive despite many adverse conditions. The Argentine military dic-
tatorship of 1976–83 was a resounding failure, and in response to the very high
price that they paid, actors that had previously been indifferent to democracy
strongly embraced it (Chapter 7, this volume). These democratic commitments
have stuck over time, perhaps because the previous dictatorship was such a
failure, and also because of the memorialization of human rights abuses under
that regime. In India, a broad, inclusive, democratic nationalism formed in re-
sponse to British colonialism (see Chapter 2, this volume). Although there have
been important challenges to that inclusive, democratic nationalism in the

Why Democracies Survive in Hard Places 229
fifty-seven years since Nehru died in 1964, the primary original organizational
carrier of that inclusive democratic nationalism, the Congress Party, has usu-
ally remained a solidly democratic actor. In South Africa, the primary opponent
to the apartheid regime, the African National Congress, became a strongly pro-
democratic actor in the 1990s. For the most part, it has remained one since then.
Finally, as Nancy Bermeo describes, in Timor-Leste, a broad democratic, inclu-
sive, national consensus also emerged in response to the widespread repudiation
of Indonesia’s brutal occupation of 1975–99.
In Benin since 2019, Georgia since 2008, India under the BJP, and Moldova
and Ukraine during many periods, ruling parties have been measurably (based
on expert surveys) less committed to liberal norms than in Argentina, India
during most of the time under the Congress Party, South Africa, and Timor-
Leste. In these contexts, democracy has broken down (Benin in 2019), never re-
ally taken hold (Georgia, Moldova, Ukraine), and eroded sharply (India under
the BJP). Indonesia is an intermediate case according to expert surveys; the
ruling parties at times (2009–14 and 2019–present) have been less committed
to liberal norms than in Argentina, India under Congress, South Africa, and
Timor-Leste, but without being as illiberal as some governing parties in Georgia,
Moldova, and Ukraine, or the BJP in India in recent years.
My argument does not pit actors’ normative preferences against institutions. To
understand why democracies survive in hard places, we also need to analyze ex-
ternal constraints that bind actors. Institutional constraints are essential because, as
James Madison (1788) famously noted in the Federalist #51, “Men are not angels.”3
For this reason, Madison argued, strong checks and balances are needed to protect
a republic. Without institutional constraints, democracy’s survival would rest solely
on actors’ willingness to accept costs to protect democracy. Although some rulers
and parties are law-and norm-respecting democrats, such people and parties do
not always hold power. Hence, the need for institutional constraints that make it
more difficult and costly to attempt to erode or overthrow democracy. Even fairly
weak institutions sometimes constrain would-be authoritarian rulers.
Perspectives that emphasize institutional constraints have put forth two po-
tentially complementary arguments. First, a strong opposition can constrain
the government and prevent executive takeovers of democracy. Enough actors
find democracy an acceptable compromise that it becomes a stable equilibrium.
Arguments that posit that democracy rests on a balance of power are intuitively
sensible, and some empirical evidence supports the idea that democracy is on
more solid ground when there is a strong partisan opposition (Pérez-Liñán et al.
2019). However, these arguments do not go very far in explaining the surprising
democratic endurance in our six cases of long-lasting democracy.4 In five of the
six cases, for extended times, government dominance could have enabled rulers
with authoritarian predilections to engage in a power grab.
Second, institutions constrain actors that would of their own accord seek
to expand their power at the expense of democracy. This was James Madison’s
focus. At first blush, it might seem that institutional arguments should not
hold much sway in democracies in hard places. Institutional checks and bal-
ances are not strong in such contexts. Nevertheless, several of the authors in
this volume (Mainwaring and Simison on Argentina, Riedl on Benin and
South Africa, Slater on Indonesia, and Varshney on India) indicate—perhaps
surprisingly—that even in these contexts of usually weak checks and balances,
institutions such as courts, federalism, and legislatures sometimes protect dem-
ocracies. Institutional constraints are not sufficient to generate regime stability
in most democracies in hard places, but they sometimes induce players to stay
in the game.
I briefly discuss two other constraints that could in principle push actors
to stick to democracy: external actors, of which the European Union and the
United States are by far the most important, and voters. The EU and the US
sometimes have encouraged transitions to democracy and prevented outright
coups, but they have little capacity to thwart incremental executive takeovers.
An older literature on voters argues that democratic voters induce regimes to
be democratic. But this literature did not make convincing arguments about
the mechanisms by which such an outcome would occur. Much of the recent
literature is skeptical that voters care deeply about democracy above and be-
yond the outputs that it produces, and hence that voters will keep democracy
intact (Graham and Svolik 2020; Svolik 2019). Even if voters care deeply about
democracy, voting is such a blunt way of conveying citizen preferences that it
is not certain that they could push politicians to abide by democratic rules of
the game.
Normative Preferences about the Political Regime
Democracies survive either because of actors’ own preferences—they embrace

democracy as the best possible political regime—or because they are constrained
to abide by democratic rules of the game. In this section, I discuss internal
constraints that stem from actors’ normative commitments. I begin with three
claims: first, that actors have very different preferences about the political re-
gime; second, that these differences in preferences are measurable; and third,
that these preferences have highly important consequences for regime outcomes.
Actors committed to democracy make it more viable; actors committed to its de-
struction or indifferent to its survival make a breakdown more likely.
Actors’ preferences regarding the political regime form a continuum ranging
from a strong normative commitment to democracy to intense hostility to it and
a preference for some kind of authoritarian regime. Imagine five points on an

underlying continuum regarding actors’ preferences about the political regime:
1. An actor has a strong normative preference for democracy. A normative

preference for democracy means that an actor is willing to forego some
instrumental benefits in order to preserve democracy. As an example,
in August 2020, Republican Senator Mitt Romney criticized President
Trump’s attacks on mail-in voting. “We should make every effort to assure
that people who want to vote get the chance to vote, and that’s even more
important than the outcome of the vote. We have got to preserve the prin-
ciple of democracy” (emphasis added).5 For Romney, democracy is more
important than winning. For actors and individuals who have a strong nor-
mative preference for democracy, building and retaining democracy and
protecting human rights are central objectives. Human rights activists and
organizations, religious leaders and churches that value human rights and
democracy, and some exceptional political leaders such as Nelson Mandela
and Jawaharlal Nehru are examples.
2. An actor has a modest normative preference for democracy. The actor
believes that democracy is the best possible political regime and is willing
to forego some instrumental advantages in order to protect democ-
racy. Most of the mainstream political parties in the advanced industrial
democracies normatively favored democracy throughout most of their
histories as mass, representative polyarchies. For the most part, these
parties practiced what Levitsky and Ziblatt (2018) call mutual tolerance
and forbearance.6 They accepted the legitimacy of the other mainstream
parties, and they refrained from constitutional hardball. In contrast, nar-
rowly self-interested instrumental actors that have no normative interest
in preserving democracy would frequently question the legitimacy of
other mainstream parties and engage in constitutional hardball on a reg-
ular basis.
3. The actor is indifferent to regime type.
4. On normative grounds, the actor rejects some aspects of liberal democ-
racy and would prefer to dismantle it. Advocates of “illiberal democracy”
such as Viktor Orbán and critics of representative democracy such as Hugo
Chávez are examples. Many populists fit into this category. Their claim to
be the true representative of the people or the nation leads them to hostility
toward democratic checks and balances and toward political opposition,
which they see as illegitimate (Galston 2018; Madrid and Weyland 2019;
Müller 2016a; Norris and Inglehart 2019; Weyland 2013, 2020).
5. The actor despises liberal democracy and would like to overthrow it and
install an authoritarian or totalitarian regime. The revolutionary left,
theocrats such as the leaders and supporters of the Iranian regime, Islamic
terrorist groups, and right-wing extremists such as the Nazis are examples.
So are personalistic autocrats who are convinced of their special, almost
innate qualities for ruling. Examples include traditional advocates of the
divine right of kings and more modern autocrats such as Rafael Trujillo
(1930–61) in the Dominican Republic. These actors are deeply (sometimes
fanatically) ideologically committed to a personalistic authoritarian or to-
talitarian regime.
Commitments to democracy are important because in weakly institutional-

ized democracies, international actors and domestic formal checks and balances
are unlikely to deter rulers who have authoritarian proclivities and significant
institutional and popular support. Just as positive normative commitments help
protect democracy, illiberal actors can help to sink it. A durable democratic equi-
librium with no democrats is extremely unlikely.
The normative preferences of presidents and prime ministers and of their
parties are particularly important; they are the prime movers of democratic pol-
itics (Pérez-Liñán et al. 2019). Executive-led suffocations of democracy hinge on
presidents, prime ministers, and governing parties who want to cripple checks
and balances and diminish opposition rights. Presidents and prime ministers
who are committed to democracy avoid such transgressions.7
Actors’ preferences about the political regime and their other program-
matic preferences are not voluntaristic individual agency. These preferences
are embedded in organizational histories, identities, donor bases, activists, and
voters. Organizational preferences are not as static as Berman (1998) suggests
in her excellent book on Social Democratic parties in Germany and Sweden,
but nor are they part of a voluntaristic world. Parties (and other actors) do not
change their programmatic profiles—including on issues related to the polit-
ical regime—overnight. There is some element of choice in these organizational
preferences, and they do change over time, but choice is bounded by the organ-
izational histories, identities, ideological commitments, donors, activists, and
bases. These are not (in my understanding) structural constraints, but they are
constraints.
Illiberalism and Democracy in Hard Places
Data from the Varieties of Party Identity and Organization (V-Party, Lührmann
et al. 2020) project are useful for indicating the degree to which the main parties
in our nine countries have embraced democratic norms. This project created an
illiberalism index based on expert surveys in 169 countries. The index is based on
four questions about party leaders’ severe personal attacks on or demonization

of political opponents; their acceptance of free and fair elections and freedom of
speech, media, assembly, and association; their acceptance of minority rights;
and their rejection of political violence. All four questions tap essential demo-
cratic norms. The scale ranges from 0 (parties that fully embrace liberal demo-
cratic values) to 1 (extremely illiberal parties). I expect that low illiberal scores
make it easier for democracy to endure, and that high illiberal scores, especially
for governing parties, make it difficult for democracy to survive in hard places.
Figure 8.1 provides information about the governing parties’ scores on V-
Party’s illiberalism scale in our nine countries at each national election. In
Argentina, South Africa, and Timor-Leste, the main parties have embraced lib-
eral democratic norms, as has the party that has governed India during most of
the period since independence—the Congress Party. The authors of the chapters
on Argentina, Timor-Leste, and India point to these normative preferences as
central to the survival of democracy in difficult circumstances.
In Argentina, consistent with Chapter 7 (this volume), two of the three gov-
erning parties since 1983, the UCR (1983–89 and 1999–2001) and PRO (2015–
19), had extremely low illiberalism scores: 0.02 to 0.03, 0.04 to 0.05, and 0.07
to 0.08, respectively. The Peronist (PJ) presidents did not always fully embrace
liberal democratic norms, but they never came close to the full-throated il-
liberalism that endangers democracy. For example, President Carlos Menem
(1989–99) packed the Supreme Court in 1990 and made extensive use of pres-
idential decrees to implement his neoliberal economic reforms, but he did not
attempt to muzzle the press, demonize the opposition, or ensconce himself in
power. As a governing party, the PJ’s illiberalism score ranged between 0.13
and 0.28 from 1989 to 1999 and between 0.21 and 0.40 from 2001 to 2015,
and it was 0.13 in 2019. Except for the PJ in the first two years of democracy
(1983–85), with an illiberalism score of 0.42, the main opposition parties
have consistently had low illiberalism scores. These scores are consistent with
Mainwaring and Simison’s argument that the most important actors have been
democracy-preserving.
Regarding South Africa, according to V-Party, the ANC initially strongly
embraced liberal democratic values, with illiberal scores of 0.15 in 1994, 0.12
in 1999, and 0.13 in 2004. Among the forces that opposed the apartheid regime,
those with a genuine commitment to democracy prevailed—the ANC faction
led by Nelson Mandela. Mandela and inner allies promoted the idea that the first
democratic government of South Africa (1994–99) would be a national unity
government that included the ruling white National Party.8 At Mandela’s in-
stigation and notwithstanding some internal critics who opposed this move,9
the ANC prioritized democratic stability over radical change. It made many
concessions to ensure that white South Africans would not feel threatened (see
Figure 8.1. Governing party illiberalism scores
Notes: The figures include all national elections (presidential and legislative) beginning with the first democratic election, and show the illiberalism score (V-Party variable v2xpa_illiberal) for the governing
party. We define the “governing party” as the party of the head of government, that is, the prime minister or the president. For each election, we show the illiberalism score for the party of the incumbent head of
government at the time of the election and for the winner in that election (i.e., if the government changes hands, then the election year will show two Illiberalism scores). For the first election of the democratic
regime, Figure 8.1 shows only the incoming governing party. If the head of government is an independent, Figure 8.1 shows the score for the largest party of their supporting coalition, measured in share of seats in
the legislature (Moldova 1998; Ukraine 1994, 1998, 2002; Benin 1991, 1999, 2007, 2011, 2015, 2019). These parties are marked with an asterisk (*). We exclude elections for which there are no illiberalism data for
the party of the head of state (Ukraine 1999, 2004, 2010; Moldova 1996, 2016; Benin 1996, 2001, 2006, 2016; Georgia 2013, 2018; Timor-Leste 2002), unless this party took part in an electoral coalition for which
there are data. In this case, Figure 8.1 includes the coalition’s data and marks them with an * (Benin 2003, India 1998, and Timor-Leste 2018). Georgia: We classify the system as presidential until and including the
2012 election, and parliamentary since the 2013 election, as a constitutional amendment significantly reduced the powers of the president in favor of the prime minister. India: For 1991, there are no illiberalism
data for the outgoing party, SJP(R), which had no allies (so I cannot use another party’s illiberalism data as proxy). For 1996: The incoming PM following the April/May elections belonged to BJP, but he lasted only
fifteen days; he resigned when it became clear that he did not have enough support to form a government. The new PM belonged to JD. Figure 8.1 includes both BJP and PD as incoming governing parties. Timor-
Leste: For 2018, there are no illiberalism data for PLP, the incoming party. Ukraine: Ukraine is a semi presidential system. Following Lucan Way’s suggestions (private communication), I classify the more powerful
executive as the president until the 2004 election, as the prime minister between the 2006 and the 2012 elections, and as the president since the 2012 election. For further details, see online Appendix 8.2.
Sources: Lührmann et al. 2020 for illiberalism scores. For sources for the head of governments and their parties, see online Appendix 8.1. For coding decisions about independent heads of government
and about elections for which V-Party did not code the party of the head of government, see online Appendix 8.2.
Figure 8.1. Continued
Chapter 4). As president (1994–99), Mandela emphasized national reconcilia-

tion and forgiveness, launching the democratic period in a promising way.
According to V-Party, the ANC took an illiberal turn in 2009 (0.41) and 2014
(0.37), but then shifted back to greater liberalism in 2019 (0.19). By V-Party’s
measure, solid support for liberal democratic principles has characterized most
of the ANC’s long stint in power. This could help explain the country’s demo-
cratic longevity despite the ANC’s long-term electoral hegemony and the weak-
ness of the legislature and courts as checks and balances.
Many democratic governments have left behind the constraints that were
negotiated during a transition. With its unassailable electoral hegemony, the
ANC had ample space to subsequently alter the terms of the transition negoti-
ations. However, the ANC remained largely wedded to an inclusive democratic
vision. It is one of the longest-established governing parties among the world’s
democracies. Its electoral hegemony makes it noteworthy that no president has
dramatically undermined checks and balances.
As Varshney argues, the Indian National Congress generated an inclusive
nationalist independence movement led by individuals—most importantly,
Gandhi and Nehru—who strongly favored an inclusive, tolerant democ-
racy (see also Tudor 2013; Tudor and Slater 2016). According to V-Party, the
Congress Party has usually embraced liberal democratic values since 1971, the
first election for which V-Party records a value for India. Its illiberalism score
was 0.29 in 1971 and 0.31 when democracy was restored in 1977—the highest
scores for the Congress Party during the nearly half-century for which V-
Party has registered one. These were the years of Indira Gandhi’s illiberal turn,
culminating in her period of authoritarian rule in 1975–77. The Congress
Party’s moderate score for 1977 reflects the rejection of Indira Gandhi’s au-
thoritarian turn. The illiberalism index for the Congress Party has since then
oscillated narrowly in a low range from 0.12 to 0.18, except for 1984, when it
increased to 0.26.
India’s first prime minister, Jawaharlal Nehru (1947–64) went to great lengths
to inspire a spirit of democratic tolerance and inclusivity. An Indian historian
commented, “For Nehru, democracy and civil liberties were absolute values,
which could not be compromised for any goal however laudable” (Mukherjee
2015, 40). Nehru’s long tenure as prime minister (1947–64) did not lead to
democracy’s success for once and for all, but it firmly planted democracy in ex-
traordinarily difficult conditions. During this period, a strong court system, a
robust federal system, a vibrant free press, a norm of vigorous electoral com-
petition, acceptance of many official languages, a majority position within the
Congress Party that India should have an inclusive and secular democracy, and
civilian control of the military emerged. Democracy is on surer footing with
these institutions and norms in place.
Except for the period from 1971 to 1977, even during periods of overwhelming
electoral dominance (1952–77, 1980–89) when it could have run roughshod
over democracy, the Congress Party has generally hewed to democratic norms.
Indira Gandhi temporarily shattered the marriage between the Congress Party
and democracy with her declaration of Emergency Rule in 1975, but in the 1977
elections voters punished her for abrogating democracy. The Congress Party
emerged from the 1975–77 period and its 1977 electoral defeat with a revitalized
commitment to democracy.
The BJP, the Hindu nationalist party, has much higher illiberal scores, con-
sistent with Varshney’s chapter. Also consistent with his argument, the BJP’s il-
liberalism represents a grave threat to democracy. During its first brief period in
government in 1996, the BJP’s illiberalism score was 0.46. During its subsequent
periods in government, its score increased from 0.43 from 1998 to 1999, to 0.53
from 1999 to 2004, to 0.58 from 2004 to 2009, and to an alarming 0.77 from 2014
to 2019, and 0.84 from 2019 to the present. Even with India’s long democratic
tradition, illiberal rulers put democracy in hard places at risk.
V-Party’s illiberal index supports Nancy Bermeo’s contention that the main
parties have been committed to democracy in Timor-Leste. Both the governing
parties and the main opposition parties have consistently scored low or moder-
ately on the illiberalism index. The National Congress for Timorese Liberation,
which headed the government from 2007 to 2017, scored 0.31 on the illiberal
index in both of its terms in office. The three largest other parties from 2007
to 2012 scored 0.32, 0.12, and 0.14, and the three largest parties after the 2012
elections scored 0.20 and 0.12, respectively. The Revolutionary Front for an
Independent Timor, which headed the minority government that took office in
2017, scored 0.20 on the illiberal index. V-Party records illiberal scores for four
other parties for the 2017 election in the very low to moderate range, ranging
from 0.11 to 0.32. A coalition, the Coalition for Change and Progress, won the
2018 election; its illiberalism score was 0.28. The other three parties for which
V-Party registered an illiberalism score ranged from 0.11 to 0.26—again, from
very low to moderate. Solid support for liberal democratic principles has helped
preserve democracy in Timor-Leste.
The argument that normative commitments can help explain the durability
of democracy under difficult circumstances works partially for Benin but with
a caveat: independent candidates won five of the country’s first seven presiden-
tial elections between 1991 and 2015. Because independents are not tethered
to a party, how parties score on the illiberalism scale is less relevant for under-
standing dynamics in the political regime.
From 1991 until 2019, the forty-two parties in Benin for which V-Party has
illiberalism scores had low to moderate scores, ranging from 0.21 to 0.48, com-
pared to a mean of 0.50 for the entire V-Party database. Illiberalism scores of
the governing party or the largest supporting party when the presidents were
independents ranged from 0.28 to 0.44 between 1991 and 2019. Until 2019, these
scores are moderate, again consistent with the argument that parties committed
to democracy can help it survive.
Two new large parties formed to support President Patrice Talon in the 2019
national assembly elections; they won 100 percent of the seats. In 2019, both
received extraordinarily high illiberal scores in V-Party (0.96 and 0.95, respec-
tively). These scores are completely consistent with the argument that if the gov-
erning parties are hostile to democracy, its prospects in difficult structural and
cultural environments are dim.
In sum, for five of our six long-standing democracies, V-Party evidence
supports the assertion that executives’ and parties’ attitudes about the political
regime correlate strongly with the fate of democracy. This argument also works
reasonably for our final case of long-standing democracy—Indonesia. V-Party’s
experts score Indonesia’s presidential parties in the early years of democracy
(1999–2004) as fairly liberal—the National Awakening Party (1999–2001) at 0.31
and the Indonesian Democratic Party of Struggle (2001–4) at 0.30. The ruling
Democratic Party (2004–14) represented a mild illiberal turn, with scores of 0.41
(2004–9) and 0.49 (2009–14). The Indonesian Democratic Party of Struggle won
the 2014 and 2019 presidential elections; V-Party’s illiberalism scores for it were
0.33 in 2014 and 0.49 in 2019. V-Party scores some of Indonesia’s other parties as
profoundly illiberal. These powerful illiberal actors would seem to make it diffi-
cult to deepen democracy and make democracy vulnerable to erosion.
V-Party’s illiberalism scores are also highly consistent with the difficulty of
establishing stable democracy in Georgia, Ukraine, and Moldova. In all three
countries, some of the governing parties and some of the other main parties
have displayed highly illiberal attitudes. Consistent with Lucan Way’s account
in Chapter 5, V-Party’s illiberalism scores of the governing parties and main
support parties (when independents were head of government) oscillate from
moderate to extreme illiberalism in Moldova and Ukraine, and from moderate
to high illiberalism in Georgia.
Table 8.1 summarizes information about V-Party’s coding of the governing
party, expectations about regime outcomes based on V-Party’s coding, and ac-
tual outcomes in our nine countries. To present the data in a synthetic way,
I focus exclusively on the governing parties; they are the prime movers in ex-
ecutive suffocations of democracy. I considered illiberalism scores of less than
0.15 extremely low, less than 0.30 low, 0.30 to 0.50 moderate, 0.50 to 0.80 high,
and above 0.80 extremely high.10 Using categorical variables for this table rather
than the actual V-Party illiberalism score makes it more efficient to summa-
rize periods during which scores fluctuated some but remained within these
categories.
Table 8.1. Illiberalism Scores of Governing Parties and Regime Outcomes
Country Elections Number V-Dem Party Probability of Actual outcome

coded by of illiberalism democratic
V-Party elections (governing survival based
coded party) on governing
during party
this time illiberalism
Argentina 1983–2007 13 Extremely Very high Survival

low or low
2009–13 3 Moderate High Survival
2015–19 3 Extremely Very high Survival
low
Benin 1991–95 2 Moderate High Survival
1999* 1 Low Very high Survival
2003*–15* 4 Moderate High Survival
2019* 1 Extremely Breakdown Electoral
high authoritarianism
(2020)
Georgia 2004 1 Moderate High Survival
2008–16 3 High Low Survival
India 1977–96 6a Low or Very high Survival
extremely low
1998 1 Moderate High Survival
1999 1 High Low Survival
2004, 2009 2 Extremely Very high Survival
low
2014–20 2 High to Breakdown Electoral
extremely authoritarianism
high
Indonesia 1999–2019 5 Moderate High Survival
Moldova 1994–98 2 Moderate High Survival
2001–9 3 Extremely Breakdown Electoral
high authoritarianism
2019 1 Extremely High Survival
low
Continued
Table 8.1. Continued
Country Elections Number V-Dem Party Probability of Actual outcome

coded by of illiberalism democratic
V-Party elections (governing survival based
coded party) on governing
during party
this time illiberalism
South 1994–2004 3 Low or Very high Survival

Africa extremely low
2019 1 Low Very high Survival
Timor- 2007–12 2 Moderate High Survival
Leste
2017–18* 2 Low Very high Survival
Ukraineb 1994* 1 Moderate High Survival
1998*– 2 High Low Competitive
2002* authoritarian in
1998
2006–7 2 High Low Competitive
authoritarian in
2011
2012–14 2 Extremely Very low Competitive
high or high authoritarian
2019 1 Low Low Survival
Note: Illiberalism scores correspond to the party of the head of government elected in each election.
An asterisk (*) indicates that the head of government is an independent and that the illiberalism data
correspond to the largest party of their supporting coalition, measured in share of seats in the legis-
lature, or that illiberalism data reflect the head of government’s electoral coalition for cases for which
V-Party does not code their party but does code the coalition. Coding of regime outcomes is based on
V-Dem version 11.1 (Coppedge et al. 2021) and follows Lührmann et al.’s (2018) coding rules.
Sources: For governing parties, see online Appendix 8.1. For illiberalism scores, Lührmann
et al. 2020.
a Following the 1996 general elections, the new BJP Prime Minister stepped down after two weeks
when he could not form a government. Table 8.1 does not include this very short-lived BJP govern-
ment. The United Front coalition then assumed office without new elections.
b Ukraine is a semi- presidential system. Following Lucan Way’s suggestions, I classify the more
powerful executive as the president until the 2004 election, as the prime minister between the 2006
and the 2012 elections, and as the president since the 2012 election. For further details, see online
Appendix 8.2.
All fifty-nine elections in which the ruling incumbent party scored low or
moderate on V-Party’s illiberalism scale were associated with the subsequent
survival of democracy during that administration according to V-Dem.11
Benin and India were democracies for a long time but then suffered severe
erosions under an independent who created highly illiberal ruling parties (Benin
after 2015) and the BJP, also a highly illiberal party (India since 2014). Moldova
and Ukraine have frequently had illiberal ruling parties, and they have not been
able to build durable democracies. Only India in 1999 and 2004 had governing
parties with illiberalism scores above 0.50 without an erosion or breakdown of
democracy—and both illiberal scores were not much above 0.50 (0.53, and 0.58,
respectively). In short, all fifty-nine illiberalism scores under 0.50 were associ-
ated with the survival of democracy according to Lührmann et al.’s (2018) coding
rules, and most scores above 0.50 were associated with sharp erosion or break-
down. These bivariate relationships do not show cause, and there is a reasonable
possibility of some coding of illiberalism based on what actually transpires in the
regime (i.e., a contamination effect). But the association between governing party
adherence to liberal principles and the survival of democracy is still striking.
What Generates and Sustains a Commitment to Democracy?

The Repudiation of the Terrible Past
In my view, actors’ normative preferences about the political regime are criti-
cally important for understanding outcomes. For example, the biggest difference
between Hungary before and after 2010 was the ascension to power of an au-
thoritarian populist who was determined to weaken democratic checks and bal-
ances and tilt the playing field toward his governing party. Thinking about actors’
attitudes about democratic norms directs us to some essential questions such as
the one I turn to in this section: what generates and sustains commitments to
democracy?
The extant literature has given little attention and few answers to this ques-
tion.12 Our volume suggests one answer for Argentina, South Africa, Timor-
Leste, and India: a widespread repudiation of the terrible past and/or a common
enemy generated broad support among organized actors and citizens for democ-
racy (see also Bermeo 1992). In Argentina, India, South Africa, and Timor-Leste,
the alternative to democracy was too unpalatable. Sustained democracy was not
an inevitable outcome, but most powerful actors and the transition leaders per-
ceived democracy as the best way to overcome the terrible past. A strong and
broad commitment to democracy did not immunize the regime from subse-
quent challenges, but the memory of the terrible past weighed heavily on polit-
ical leaders and in favor of democracy.
In Argentina, the failures of the dictatorship of 1976–83 were many and pro-
found, as Mainwaring and Simison argue in Chapter 7. Economic performance
was dismal. When the regime’s support faltered because of economic misman-
agement, the government waged a reckless and disastrous war against the United
Kingdom by invading the Falkland/Malvinas Islands in 1982. The cost in lives
(about 650 Argentine soldiers died) and economic ruin was enormous. The lo-
gistical preparation was dismal, leading to appalling conditions for the mostly
young Argentine soldiers who invaded; without adequate food, boots, and coats,
they starved and froze until the generals capitulated. To garner support for the
war, even as the Argentine military was being routed, right up until the day they
surrendered, with a massive propaganda effort, the generals proclaimed that
they were winning the war. Their brazen lies reinforced the effects of the bellicose
disaster, discrediting the military as a bunch of incompetents and liars. After
the transition, in 1985–86, the world’s most publicized human rights trials since
Nuremburg in 1945–46 galvanized public opinion against the human rights
atrocities committed by the military.
Mainwaring and Simison argue that the broad repudiation of the dicta-
torship of 1976–83 generated new commitments to democracy and a rejec-
tion of extremist positions (see also Ollier 2009). The Peronist Party and the
labor unions, which had never been committed to democracy, became ardent
supporters; some leaders risked their lives to protect democracy during the mil-
itary rebellions of the 1980s and 1990. The revolutionary left was vanquished,
and a few of its participants became intellectual leaders celebrating the virtues
of democracy. Business associations, which had embraced the coups of 1955,
1962, 1966, and 1973, now recognized that untethered dictatorships could be
very destructive; they, too, rejected the authoritarian past. Since 1983, a strong
human rights movements, successful prosecutions of the human rights abuses
committed during the military dictatorship, and a powerful memorialization of
the past abuses have helped maintain aversion to authoritarian leaders.
Nancy Bermeo argues that armed conflict in Timor-Leste paved the way to
a durable democracy. The primary armed conflict pitted different groups in
Timor-Leste against the Indonesian invaders and intruders, thereby creating co-
hesion among previously fractious groups against the hated enemy. The inde-
pendence war “strengthened an inclusive national identity” and “gave rise to a
series of inclusive institutions (and mindsets)” (Chapter 6, this volume). First, a
sanguinary civil war broke out in 1975 between warring factions within Timor-
Leste. Then the Indonesian military invaded, leading to a twenty-four-year-long
brutal occupation (1975–99) that generated widespread resentment toward the
occupying force and a newfound sense of inclusive nationalism. If before the
occupation the East Timorese had conflictual understandings of national iden-
tity, the Indonesian occupation created a sense of a common enemy that in turn
created a cohesive inclusive nationalism. Bermeo notes that in response to the

need for a cohesive national project forged in opposition to Indonesia, the char-
ismatic party leader Xanana Gusmão renounced his and his party’s Marxist rev-
olutionary path and instead embraced democracy as a project that would bring
Timor-Leste together. The war of independence also had a professionalizing ef-
fect on the military.
In South Africa, as Riedl argues, the leadership of the African National
Congress, the vast majority of the Black population, and most of the interna-
tional community widely abhorred apartheid. The South African regime’s ex-
treme racial discrimination made it a worldwide pariah.
Just as the Argentine military dictatorship, the South African apartheid re-
gime, and the Indonesian invaders of Timor-Leste formed hated enemies that
helped forge a widespread democracy movement, so did the British colonialists
in India. To help defeat the British and achieve independence, by the 1930s, the
Indian National Congress mobilized scores of millions of Indians in opposition
to colonial rule and on behalf of independence. The broad, inclusive, tolerant na-
ture of the movement also entailed a deep commitment to inclusive democracy
on the part of Gandhi and Nehru, as well as many other Congress Party leaders.
Gandhi and Nehru’s vision for this inclusive democracy shaped India’s con-
stitution and the Congress Party, generating momentum for democracy that
lasted for the first two decades (Tudor 2013; Tudor and Slater 2016). Congress’s
commitments helped lead to a constitution that enshrined the principle that de-
mocracy had to be for all Indians, including the poorest, all religious groups, and
the lower castes (see Chapter 2, this volume). They set the tone for the Congress
Party and for India’s political system during the first decades after independence
in 1947.
In Argentina, India, South Africa, and Timor-Leste, because of the wide-
spread repudiation of the terrible past, no electorally successful party reclaimed
the legacy of the previous regime. The only very partial and ephemeral exception
to this rule was the National Party in South Africa, which won 20.4 percent of the
vote in the founding election of 1994, but it was still dwarfed by the ANC, which
won 62.7 percent. The party rebranded itself the New National Party in 1997
to create some distance from its apartheid past, but its vote share collapsed to
6.9 percent in 1999 and 1.7 percent in 2004 before the party disbanded in 2005.
The absence of parties to reclaim the legacy of the authoritarian or colonial past in
these four cases underscores the thorough rejection of the previous dictatorships.
No authoritarian successor party can mobilize authoritarian sentiments. This
absence is a dramatic contrast to the many democracies in the world that have
thriving authoritarian successor parties (Loxton and Mainwaring 2018).
In India, Timor-Leste, South Africa, and Argentina, the commitment to de-
mocracy evinced by the governing party or parties and the main opposition
parties was enormously helpful. Commitment to democracy enabled these

regimes to survive in contexts of great adversity. A critical question for future re-
search is how to keep alive normative commitments to democracy.
These four cases were not unique in generating commitments to democracy
that endured for some time based on a repudiation of the terrible past. In Chile,
after the left experienced severe repression during the Pinochet dictatorship
(1973–90), it became firmly committed to democracy (Roberts 1998; Walker
1990). In similar fashion, most of the Brazilian, Uruguayan, and Spanish left be-
came converts to democracy after the dictatorships of 1964–85, 1973–85, and
1939–75, respectively. A traumatic experience such as a repressive dictatorship
(Argentina, Timor-Leste, and South Africa), a foreign invasion and civil war
(Timor-Leste), or an extremely racist regime that oppressed the great majority of
the population (South Africa) does not inexorably lead individuals and actors to
value democracy—but it did so in these cases for an extended time.
Democracy as an Equilibrium of Narrowly

Self-Interested Actors
Some explanations posit that democracy endures because self-interested actors

reluctantly conclude that democracy is their best option because of a power equi-
librium, or because institutions constrain the military and the executive from
overthrowing the regime. These approaches focus exclusively on constraints ex-
ternal to actors and regard actors as narrowly self-interested. They do not con-
sider actors’ normative commitments part of what makes democracy viable.
These approaches often help explain democratic stabilization, but I am skeptical
that they are sufficient.
By nakedly or narrowly self-interested, I mean something along the lines
that Przeworski (2019, 19–20) wrote: “The dream of all politicians is to conquer
power and to hold on to it forever. It is unreasonable to expect that competing
parties would abstain from doing whatever they can do to enhance their electoral
advantage.”13 I discuss two approaches that view democracy as an equilibrium
among narrowly self-interested actors. One posits that democracy is an equilib-
rium among actors, none of which can overpower the others. These actors accept
democracy because they can’t overpower the others, and because democracy
is a tenable compromise that gives them enough of what they want. The other
sees democracy as an equilibrium in which formal institutions—the legislature,
the courts, and other mechanisms of accountability—prevent rulers from over-
throwing democracy. My own view is that scholars must seriously consider both
possibilities, but that they do not go far toward explaining democratic stability in
hard places—at least for the cases studied in detail in this book.
A Large Partisan Opposition
One plausible answer to what sustains democracies is a large partisan opposition;

the opposition controls enough levers of power (such as seats in the national as-
sembly, governorships, etc.) to prevent the government from encroaching deeply
on democracy or from enacting radical measures that would deeply threaten
established interests, potentially leading to the formation of a coup coalition.
Government majorities present two potential threats to democracy. On one
hand, they facilitate executive takeovers. A minority opposition usually cannot
block political reforms that enhance executive power, bolster the position of
the governing party, allow the executive to capture supposedly independent
institutions, and marginalize the opposition. On the other, because governments
with majorities might not face meaningful checks and balances, it might be
easier for them to pursue potentially destabilizing radical agendas.
The focus on partisan checks on the government has a long pedigree.
Przeworski (1991,10–50; 2019) advanced a partisan balancing argument. He
wrote (2019, 177) that “checks and balances do not operate effectively when
different powers of the government are controlled by the same party.” Rustow
(1970) and other scholars have advocated a balance-of-power argument to ex-
plain pacted democratic transitions—although Rustow explicitly did not extend
his argument to explaining democratic endurance.14 Lijphart (1977) advocated
power-sharing arrangements including multipartism, especially in contexts
of democracy in plural societies, as a way of balancing power. In this volume,
Riedl’s chapter on Benin and South Africa advocates a balance-of-power argu-
ment as a way of understanding democratic stability.
Little empirical work has systematically tested arguments about the impact
of a balance of power on the survival of democracy. One exception is Pérez-
Liñán et al. (2019), who provide compelling evidence that a sizable opposition
was helpful for democratic longevity in Latin America from 1925 to 2016. When
presidents controlled the legislature and courts, democratic breakdowns were
more likely.
Our volume offers surprisingly thin evidence to support the idea that opposi-
tion majorities help protect democracy in hard places. Figure 8.2 shows the per-
centage of assembly seats controlled by the governing party and by the governing
coalition for our nine countries. To operationalize whether a party is part of the
government, I used the parties that hold cabinet portfolios. The dataset on cab-
inet composition is based on data in July of every year, and it extends until at least
2016 and for some countries later. Figure 8.2 includes data for the upper chamber
only for Argentina, the only case in this volume that has symmetrical bicamer-
alism. In Georgia, cabinets have included only the governing party, so Figure 8.2
has no second line for this country for the seat share of the governing coalition.
Figure 8.2. Percentage of national assembly seats controlled by the governing party and coalition
Note: When the head of government is an independent, the seat share of the governing party is 0%. When cabinet periods do not begin on the first day of a month, we
assigned them to the next month (e.g., the period beginning on 10/29/84 is set as beginning in 11/1984).
Sources: For a complete list of references, see online Appendix 8.1.
Figure 8.2. Continued
Among our nine cases, none has consistently exhibited a strong partisan oppo-
sition. In Benin, the governing coalition enjoyed a modest majority (55 percent)
in 1996–98, 2003–6 (58 percent), and 2011–15 (52 percent) and a commanding
majority in 2006–7 (72 percent), 2010–11 (66 percent), and 2017–19, at which
point the president’s attacks on democracy were already under way (81 percent
from 2017–19, then 100 percent). Paradoxically, one of the smallest seat shares
controlled by the president’s coalition, 29 percent of the seats, occurred during
the first eighteen months of Patrice Talon’s presidency, in 2016–17. Talon’s super-
majorities from October 2017 on were a product of his aggressive attacks on de-
mocracy rather than initially being a necessary or even a facilitating condition
for democratic erosion.
In Indonesia, the president’s coalition has consistently had sizable majorities
in the national assembly, reaching 93 to 95 percent in the early democratic period
(1999–2004) and never dipping below 73 percent until 2014, when it dropped to
60 percent. Slater (2018b) referred to this phenomenon of oversized coalitions as
“promiscuous power-sharing.” Presidents form grand coalitions, and few parties
end up in the opposition. Thus, there is little legislative opposition to block po-
tential presidential encroachments.
In South Africa, India, and Georgia, the governing party has sometimes had
large majorities that would make it easier for a president or prime minister to en-
gage in executive encroachments. In all three cases, at times, the government was
so dominant that the opposition would have had a difficult time blocking executive
encroachments. Despite enjoying large legislative supermajorities, the presidents and
prime ministers during these periods generally respected the rules of democracy.
In South Africa, the ANC has consistently won a large majority of seats.
Consistent landslides have not led to dramatic executive efforts to erode de-
mocracy. The South African case violates the notion that democracy rests on the
losers’ willingness to accept results today because they might win tomorrow; no
contender has ever had a reasonable chance of defeating the ANC.
Since redemocratization in 1977, India has several times had weak partisan
opposition in the dominant lower chamber (Lok Sabha). In the first decades
of India’s democracy, the Congress Party was overwhelmingly dominant. It
won 364 of 489 seats in the 1951–52 parliamentary elections; such dominance
could easily have facilitated an executive takeover of democracy. In the 1980s,
the Congress Party again won huge majorities of seats. In 1980, it won 353 of
529 seats (67 percent), and in 1984 it won 404 of 514 seats (79 percent), its most
dominant election ever; the second biggest party won only 30 seats. The Indian
parliament was rarely a check on the power of the government.15 Since the 2019
elections, the BJP has had a solid majority (56 percent) in parliament, facilitating
the attacks on liberal rights that Varshney details in Chapter 2.
In Argentina, the data in Figure 8.2 about the seat share of the governing party
and the governing coalition understate the degree to which Peronism has dom-
inated elections and the political system since 2001. Peronists have won four of
five presidential elections, three times in landslides (2003, 2007, 2011). In 2003,
a fractured Peronist field won 61 percent of the valid vote distributed among
three different PJ candidates. The leading non-Peronist won only 16.4 percent of
the vote. In 2007, Cristina Fernández de Kirchner won 45 percent of the vote, a
22 percent margin over her closest competitor, and in 2011 she won 54 percent
of the vote, a 37 percent margin over the runner-up. In 2019, Alberto Fernández
won by an 8 percent margin. Peronism has also usually dominated the national
congress. Moreover, some of the nominally non-Peronist parties in an increas-
ingly fragmented party system fairly consistently align with the president.
Although the national congress has occasionally blocked presidential initiatives,
it would be far-fetched to attribute the survival of democracy in Argentina to a
strong partisan opposition.
In Timor-Leste, oversized majority governing coalitions have been the rule
rather than the exception. At the extreme, in 2015–17, the president’s cabinet in-
cluded parties that occupied all of the seats in the national assembly. In 2017–18,
the presidential coalition controlled 92 percent of the seats.
In Moldova, the governing coalitions under both democracy and authoritar-
ianism between 1994 and 2005 and again since 2019 controlled large majorities
(more than 60 percent) in the national legislature. The same was true in Ukraine
from 2006 to 2010 and from 2014 to 2016.
In sum, although a partisan balance of power is an asset for democratic sta-
bility, in all of our cases, governing coalitions have had large majorities during
extended periods. The stability of democracy in our nine cases has not rested on
the opposition’s capacity to block the executive.
Moreover, in some cases (although not among our nine until perhaps Benin
since 2016), rulers with authoritarian ambitions have undermined democ-
racy in part because they faced partisan shackles in the national legislature.
Well-known cases include Alberto Fujimori in Peru (1990–2000), who closed
the congress and courts in April 1992 to eliminate legislative constraints that
blocked him from taking actions (Kenney 2004); Hugo Chávez in Venezuela
(1999–2013), who almost immediately and with dubious legality convened
a constitutional congress to circumvent the recently elected congress, whose
partisan composition was not favorable to him; and Rafael Correa in Ecuador
(2007–17), who did the same for the same reason. Correa created a new
party when he ran for president, but his party did not have any congressional
candidates, so the legislature had no Correa copartisans when he assumed of-
fice in January 2007.
Democracy as an Intertemporal Bargain

A variant of the partisan balance argument posits that democracies are stable
because of an intertemporal bargain. Müller (2016a, 254) expressed this idea
clearly: “Democracies rely on the idea that losers in elections have every
reason to continue playing the political game, because they might form a
majority the next time” (see also Przeworski 1991, 10–50). Rulers who were
initially freely and fairly elected but then presided over executive takeovers
manifest a flaw in this logic. These rulers rigged the system in their own
favor, thus simultaneously reaping more of the system’s benefits today (e.g.,
through patronage and cronyism) and diminishing, perhaps radically so, the
likelihood of losing tomorrow. If the winners of elections today did “what-
ever they can to enhance their electoral advantage” (Przeworski 2019), they
should seize their temporary electoral advantage and cripple the opposition.
This does occur, but when it is a regular part of politics, it signals that a de-
mocracy is already ill.
In sum, in contexts of weak institutions, illiberal rulers can sometimes ma-
neuver around partisan constraints. Rules and procedures that should in prin-
ciple bind these heads of government do not. If I were designing electoral rules
to reduce the likelihood of a breakdown of a generic democracy in a hard place,
other things equal, I would recommend rules that promote multipartism, aim
to create a partisan balance of power, and make it difficult for one party to com-
mand majorities. But in the cases in this volume, democratic survival has not
rested on a consistently strong partisan opposition.
Many accounts of pacted democratic transitions have convincingly argued
that they involved a power balance among self-interested actors. Outgoing re-
gime elites won some protections for their interests, and incoming democrats
won democracy (Riedl’s chapter, this volume; Rustow 1970; Ziblatt 2017). Both
sides won some advantages by agreeing to the democratic transition. But the
fact that pacted transitions constitute an equilibrium of narrowly self-interested
actors does not explain why it remains in the interests of the government to not
defect at subsequent moments. The initial balance of power at the time of the
democratic transition does not get frozen in time. An account of democracy as
a balance of power among self-interested actors must explain not only the ini-
tial transition but also why the actors continue to find democracy an acceptable
compromise.16
Implicitly, the argument that democracy is a stable equilibrium because nar-
rowly self-interested actors find it in their interests to preserve the regime rests
on the assumption that actors pay a fairly high price for defecting from democ-
racy, or that democracy consistently yields substantial payoffs to all powerful
players. If narrowly self-interested rulers do not pay a price for encroachments,
they should regularly seize opportunities to grab advantages and reduce the
likelihood of having to relinquish power. This kind of behavior does occur, but
when it is the norm, democracy rests on precarious bases.
Legislatures and Courts as Protectors of Democracy
Several chapters in this volume suggest that while formal institutions are far from
sufficient, they do help protect democracies. Riedl on South Africa and Benin,
Slater on Indonesia, and Varshney on India convincingly argue that formal
institutions have helped protect these democracies. This is perhaps somewhat
surprising because as Brinks et al. (2019), Levitsky and Murillo (2005, 2014), and
O’Donnell (1993, 1994) have suggested, formal institutions are almost always
much weaker in democracies in hard places. They do not consistently provide
robust protection for democracy.
In several of our nine countries, institutional arrangements have protected
democracy in at least two ways. First, some institutions have promoted power
sharing and inclusion; they are contra-majoritarian. They have given a broader
range of actors and interests a stake in the game, and therefore, they dampen
incentives to overthrow democracy. In Benin, proportional representation for
the national assembly helped ensure the representation of different ethnicities
and averted “ethno- regional political dominance” (Riedl, Chapter 4, this
volume). Varshney on India and Riedl on Benin and South Africa argue that
constitutional and institutional design helped bring about democratic stability.
Quasi-federalism in India has allowed groups that are minorities at the country
level to govern at the state level, thus helping to integrate a hugely diverse country
(Stepan et al. 2011; Varshney 2013, chapter 6).17 Federalism thereby facilitated
the peaceful articulation of some ethnic, religious, and linguistic demands in a
context of extraordinary diversity. Varshney points to another way in which fed-
eralism has contributed to the survival of democracy in India: the parties that
rule at the state level but in the opposition at the country level push back against
executive encroachments.
In South Africa, the transitional compromise resulted in a constitution that
protected whites’ economic interests and fostered moderation; both decisions
were favorable to the establishment and preservation of democracy. Riedl argues
that federalism in South Africa gave the Zulu and white minorities control over
at least one of the nine provinces and protected elite economic interests. This
arrangement helped create buy-in among these groups and thereby helped
pave the way for democracy. Likewise, an asymmetrical quasi-federal arrange-
ment in Indonesia has helped balance what Slater (Chapter 3) calls inclusive
nationalism with some regional ethnonationalist demands (Bertrand 2007).18
Federalism in South Africa and quasi-federalism in Indonesia and India have
helped democracy. In large countries with as much “riotous diversity,” to use Dan
Slater’s felicitous expression, as India and Indonesia have, federalism has been an
asset for democratic stability.
Contra-majoritarian, consensus-building, inclusive institutions are generally
favorable to preserving democracy (Ginsberg and Huq 2018, 164–204; Lijphart
1977; Pérez-Liñán et al. 2019; Reynolds 2002) in contexts of politically salient
ethnic, national, or religious cleavages—a huge challenge in India and Indonesia,
among many other democracies in hard places. They divide and limit power.
They encourage accommodation and efforts to build a consensus across different
groups, and they give a broader range of actors access to parts of state power.
A second way in which institutions have helped democracy is that legislatures
and courts have sometimes limited executive power. The Argentine case provides
some examples. When Carlos Menem attempted to run for a third consecutive
term for president in 1999, the Supreme Court’s public lack of support made it
impossible for him to go ahead (Helmke 2005, 135–41).
However, institutions don’t offer a general explanation for the longevity of de-
mocracy in hard places because the institutions that are designed to check exec-
utive power are rarely strong. Legal systems are vulnerable to executive influence
and control. In contexts of weak institutions, popular rulers who want to run
roughshod over democratic checks and balances often can find ways to do so.
This is the peril now facing democracy in Benin and India—and a common peril
of democracies with relatively weak institutions.
As an example of institutional fragility in the face of a determined executive
assault, Patrice Talon, Benin’s wealthiest businessman, won the 2016 presidential
election, running as an independent. When he assumed office, little indicated
that he would quickly set about undermining Benin’s democracy. Unlike many
populists, he did not rail against democracy during his campaign. The national
assembly, which had been elected in 2015, was fragmented along party lines.
Talon initially had a minority in the National Assembly, and the largest party
was in the opposition. As a result, he was seemingly poorly positioned to capture
institutions. He initially pledged that he would not seek reelection after his five-
year term.
This innocuous beginning soon gave rise, however, to presidential incursions
against democracy. Talon attacked the independence of the courts and the media
and began to target political opponents. In February 2018, he appointed his per-
sonal attorney as head of the Constitutional Court. Before this move, Benin’s
Constitutional Court had a history of judicial independence. With his personal
attorney as head of the Constitutional Court, this independence quickly eroded.
Before Talon’s presidency, Benin had a strong history of free and fair elections.
The Electoral Integrity Project scored Benin as having the second freest and
fairest elections in Africa from 2012 to 2018, marginally behind Cape Verde (70
points for Benin and 71 for Cape Verde on a scale of 0 to 100), and ahead of the
United States (61 points on the 100-point scale) (Norris and Grömping 2019).
This changed dramatically after Talon captured the courts and the national as-
sembly. In 2018, at Talon’s behest, the national assembly passed a new, highly
restrictive electoral law that regulated the 2019 elections for the new assembly.
The Constitutional Court also required that parties obtain a “certificate of con-
formity” to be eligible to run for the elections. Only two parties, both of which
backed Talon, qualified to run. The new rules “essentially made it impossible for
anyone other than Mr. Talon’s supporters to run for office” (Maslin Nir 2019).
Intimidated opposition observers canceled plans to monitor the elections.
When the opposition boycotted the elections because of the highly skewed
playing field, turnout plummeted to about one-fourth of eligible voters, down
from 68 percent in 2015. One hundred percent of the national assembly seated in
2019 supports Talon; there is no opposition. The government also greatly stepped
up harassment of the opposition. Former president Thomas Boni Yayi (2006–16)
was under house arrest for fifty-two days after the April elections, and then he fled
Benin. Talon’s main rival in the 2016 presidential election, Lionel Zinsou, was
banned from running for office for five years. In 2018, the third-place finisher in
the 2016 presidential election, Sebastian Ajavon, was condemned to twenty years
in prison. The High Authority for Audiovisual Media and Communication shut
down some opposition media. In 2017, a new digital media law facilitated sub-
sequent harassment and legal cases against opposition and independent media.
The police and military have violently suppressed opposition protests.
It seems stunning how quickly and easily Talon was able to undo the previous
twenty-five years of democracy. The institutions that had been in place during
Benin’s democracy from 1991 to 2016 were intact when Talon took office, but
a president determined to subvert democracy was able to do so in short order.
As Ashutosh Varshney writes in Chapter 2 of this volume, India under Prime
Minister Narendra Modi is another example of a leader with an authoritarian
orientation who has been able to carry out a project of eroding democracy de-
spite the country’s long democratic tradition. Democratic institutions that once
appeared to be solid have crumbled under pressure.
Institutional Weakness: Legislatures
Legislatures and courts in our nine cases are not strong. Presidents and prime
ministers have an arsenal of weapons that often enable them to circumvent
constraints that legislatures and courts should constitutionally be able to impose.
Legislatures are supposed to provide a counterbalance to presidents and prime
ministers, and they sometimes do—although less so if the governing party has a
majority in the national legislature. While legislatures sometimes constrain ex-

ecutive power,19 in the cases in this volume, there is limited evidence that they
have fostered the longevity of democracy by successfully pushing back against
executive power grabs.
Legislatures often fail to constrain rulers. As noted in the previous section,
in our nine countries, heads of government have often enjoyed supermajorities
in the national assemblies. In such situations, legislative checks generally do
not work in weakly institutionalized democracies. Moreover, even when heads
of government do not enjoy solid majorities in the national assembly, in most
third-wave democracies, they have powerful tools at their disposal that could
enable them to push beyond the limits of democracy. Many presidential systems
grant strong constitutional powers to the executive, giving presidents more
leeway than they have in the US (Cox and Morgenstern 2001). Most democrat-
ically elected presidents have more extensive lawmaking and decree capacity
than the US president (Carey and Shugart 1998; Figueiredo and Limongi 1999).
These features make it easier for the executive to capture institutions and run
over democracy. Other things equal, prime ministers who have majorities in the
national assembly might be even more likely to be dominant politically because
parliamentary systems have a fused election and survival of the head of govern-
ment. Therefore, they usually do not contend with a legislature whose purpose
is, in part, to serve as a counterbalance to the executive.
Chernykh et al. (2017) constructed an index of legislative powers (the
Weighted Legislative Powers Score) for 158 countries based on expert surveys
for the years around 2005–7. Scores are based on legislatures’ legal and consti-
tutional authority, and they range from 0 (an extremely weak legislature) to 5.93
(Germany). Only one of our nine legislatures, Moldova, has strong constitu-
tional powers. The other eight countries featured in our volume scored in the
second and third quartiles (see Table 8.2), although the sample included many
authoritarian regimes with weak legislatures. In many third-wave democracies,
presidents can circumvent legislatures that they view as obstructionist.
In principle, if the upper chamber of the national assembly is powerful, bi-
cameralism might serve as a protection against power-hungry executives eager
to undermine democracy. However, of our nine cases, only Argentina has a pow-
erful upper chamber, and its greatly malapportioned Senate has hampered rather
than bolstered democracy. Argentina’s federal system has rarely been a road-
block for presidents who wanted to expand their power. On the contrary, it has
benefited authoritarian provincial political bosses, with negative consequences
for democracy at the country level (Gervasoni 2018).
In sum, legislatures sometimes limit executive encroachments. But as the
above discussion of Benin suggests, and as cases such as Venezuela in 1999 and
Ecuador in 2007 also show, even majority opposition legislatures do not always
block executive takeovers. Legislative constraints are not sufficient to protect
Table 8.2. Weighted Legislative Powers Scores for

Nine Countries
Weighted Weighted
legislative legislative powers
powers score rank (out of 158)
Argentina 3.60 76
Benin 4.13 58
Georgia 4.35 50
India 4.51 47
Indonesia 4.25 47
Moldova 5.29 21
South Africa 4.57 43
Timor-Leste 3.43 82
Ukraine 4.08 62
Source: Chernykh et al. 2017.
democracy, and in some contexts, especially when the executive’s party also
controls the legislature, they do not protect it at all. In weakly institutionalized
democracies, presidents and prime ministers typically have an arsenal of tools
that enable them to be dominant players.
Institutional Weakness: Courts
A similar story is true for courts. In an era of increasing judicialization of

democratic politics, courts have become meaningful actors in many third-
wave democracies. They sometimes issue rulings that limit or prevent exec-
utive encroachments (see Varshney, Chapter 2 in this volume), and in a small
number of third-wave democracies they are powerful actors that can limit ex-
ecutive encroachments. However, in most third-wave democracies, courts are
not strong, independent actors. Presidents and prime ministers with hegemonic
ambitions are often able to violate or quietly ignore court rulings and constitu-
tional norms. They may be able to intimidate or offer rewards to judges to step
down and replace them with more pliant individuals. They might be able to pack
the courts. In most poor democracies, courts are underfunded, understaffed,
and ill prepared to take on presidents and prime ministers with authoritarian
instincts.20 The judiciary struggles to achieve independence in relation to the ex-
ecutive and other powerful actors.
Expert surveys show a perception of relatively weak courts in most of our nine
cases. Four V-Dem questions focus on judicial capacity to check the government.
Question 3.7.0.11 of V-Dem Version 10 (February 2020) asks, “When the high
court . . . is ruling in cases that are salient to the government, how often would you
say that it makes decisions that merely reflect government wishes regardless of its
sincere view of the legal record?” (Item v2juhcind). Question 3.7.0.12 asks, “When
judges not on the high court are ruling in cases that are salient to the government,
how often would you say that it makes decisions that merely reflect government
wishes regardless of their sincere view of the legal record?” (Item v2juncind).
Scores range from 0 (“always”) to 4 (“never”).21 Columns 2 and 3 of Table 8.3 show
V-Dem’s linearized original scale posterior estimate, which treats the ordinal
values on the original scale as if they were linear (Coppedge et al. 2020b, 25–26).
According to V-Dem’s experts, the high courts in South Africa do not often
issue rulings simply to accommodate the government. But the coders judge
that the other eight countries have pliant courts that with some frequency make
“decisions that merely reflect government wishes regardless of their sincere
view of the legal record.” In Argentina, Benin, Georgia, Moldova, Timor-Leste,
and Ukraine, the point estimate in Table 8.3 is closer to a judgment that courts’
Table 8.3. Expert Perception of Judicial Capacity to Constrain Governments,

V-Dem (2019)
High court Lower court Government Compliance

independence independence compliance with other
(3.7.0.11) (3.7.0.12) with high courts
court (3.7.0.14)
(3.7.0.13)
Argentina 2.35 2.36 3.04 2.99

Benin 2.22 2.33 2.84 2.37
Georgia 2.11 1.30 3.05 2.66
India 2.72 2.13 2.84 2.82
Indonesia 2.73 2.10 3.22 2.96
Moldova 2.60 2.74 2.66 2.63
South Africa 3.40 3.05 3.09 3.03
Timor-Leste 2.07 2.45 1.77 2.48
Ukraine 1.64 1.72 1.58 1.27
Denmark 3.67 3.68 3.63 3.43
Note: The V-Dem question number is in parentheses in the top row.

Source: Coppedge et al. 2020b.
decisions bow to government preferences about half the time “regardless of their
sincere view of the legal record.” The surveys judged the lower courts in Ukraine
and Georgia to be particularly pliant. The final row, in italics, shows perceptions
for Denmark, a country with strong legal institutions, to provide one benchmark.
Question 3.7.0.13 is “How often would you say the government complies
with important decisions by the high court with which it disagrees?” (Item
v2juhccomp). Question 3.7.0.14 is “How often would you say the government
complies with important decisions by other courts with which it disagrees?”
Scores range from 0 (“never”) to 4 (“always”) (Item v2jucomp).22
As the final two columns of Table 8.3 show, no country’s score is much above
3.00 for either question. A score of 3.00 indicates a perception that the govern-
ment “usually” “complies with important (judicial) decisions . . . with which it
disagrees.” Obviously, it is very problematic for a democracy if a government
often does not comply with important judicial decisions.
The World Bank Governance Indicators include two measures that are rele-
vant to understanding perceptions of institutional capacity to check rulers with
authoritarian proclivities: control of corruption and rule of law. Where corrup-
tion is pervasive and where the rule of law is weak, it is presumably easier for
would-be authoritarians to run over institutions and degrade democracy. Hence,
democracy should be on firmer footing with solid scores for perceptions of con-
trol of corruption and rule of law.
Table 8.4 shows how our nine countries scores on these two measures. Scores
are the number of standard deviations above and below the world mean in a
Table 8.4. World Bank Governance Indicators: Perceptions of Control

of Corruption and Rule of Law in Nine Hard Cases, 2019
Control of corruption Rule of law
Argentina –0.07 –0.43

Benin –0.32 –0.66
Georgia 0.67 0.31
India –0.23 –0.03
Indonesia –0.42 –0.34
Moldova –0.62 –0.37
South Africa 0.08 –0.08
Timor-Leste –0.38 –1.11
Ukraine –0.71 –0.70
Source: Worldwide Governance Indicators, n.d.

given year, and they are based on expert and public opinion surveys. With the
surprising exception of Georgia, none of the eight countries scores well on these
indicators. Benin and Timor-Leste (and Moldova and Ukraine, which are not
cases of long-lasting democracy) have poor perceptions for control of corrup-
tion and rule of law, making the longevity of their democracies all the more
notable.
The ease with which executives with authoritarian predilections have
steamrolled courts in weakly institutionalized democracies makes us skeptical
that the reason for democratic resilience in our cases has much to do with a vig-
orous and independent legal system. None of the authors of the cases suggests
that the court system has been a major asset for democratic longevity.
In sum, in almost all democracies in hard places, institutional weakness is an
ongoing problem. Counting on institutions alone, or even in combination with a
partisan balance, to protect democracy is a poor bet.
Effective Governance and Durable Democracy
An obvious hypothesis for why democracy might endure in hard places is that
governments function well despite the difficult environmental circumstances.
Some of the extensive literature on economic performance and democratic en-
durance supports this argument, but the findings in this literature are mixed.
In Table 1.1 in the introduction to this volume, Models 1 and 4 show a weakly
statistically significant correlation (p < .10) between poor growth performance
and democratic breakdown; in the other four models, economic growth is not
significant.23
The authors in this volume do not invoke solid government performance
as a reason for why democracy endured for a considerable time in difficult
circumstances. In fact, it is difficult or impossible to attribute democratic endur-
ance in most of our cases of long-standing democracy to good government per-
formance. Table 8.5 shows one common measure of government performance,
per capita GDP growth under democracy. Among our six long-standing democ-
racies, India and Indonesia registered strong average growth rates. The average
growth record in Argentina and South Africa has been poor, and in Benin, it has
been mediocre. Table 8.5 also provides the data for the three post-Soviet cases.
Table 8.5 also shows the World Governance Indicator for Government
Effectiveness for 2019. Scores are the number of standard deviations above or
below the world mean of 0 in a given year. Except for Georgia and South Africa,
the World Governance Indicator registers average (Argentina, India, Indonesia)
or fairly poor (Benin, Moldova, Timor-Leste, and Ukraine) perceptions of gov-
ernment effectiveness. Effective governance certainly helps protect democracy,
Table 8.5. Perception of Government Effectiveness and Per Capita GDP Growth
in Nine Hard Cases
Country Period Average GDP per Perception of

capita growth government
effectiveness (2019)
Argentina 1983–2019 0.97% –0.09

Benin 1991–2019 1.45% –0.44
Georgia 2012–2019 4.51% 0.83
India 1977–2019 4.00% 0.17
Indonesia 1999–2019 3.69% 0.18
Moldova 1994–2019 3.61%a –0.38
South Africa 1994–2019 1.12% 0.37
Timor-Leste 2002–2019 1.82% –0.88
Ukraine 1994–2019 0.90% –0.30
Sources: GDP per capita growth, World Development Indicators, n.d.; government effectiveness,
Worldwide Governance Indicators, n.d.
a There are no data for 1994 and 1995.
but it does not explain why democracy has survived for a considerable time in
most of our cases.
International Constraints: The EU, the US, and the OAS
Another possible reason why democracies might survive is international sup-

port for democracy and pressures against individuals and groups that attempt
to erode or overthrow it. Two generations of research have shown that inter-
national actors and influences affect prospects for democracy (Brinks and
Coppedge 2006; Gleditsch 2002; Gleditsch and Ward 2006; Levitsky and Way
2010; Mainwaring and Pérez-Liñán 2013; Pevehouse 2005; Vachudova 2005,
2010; Whitehead 1986).
In the 1970s and 1980s, EU support for democratization in Greece, Portugal,
and Spain helped nurture democratic consolidation (Pridham 1991; Whitehead
1986). The 1990s generated optimism that international actors and influences
could help sustain democracies. The European Union embarked on the most
ambitious democracy promotion ever. With EU support, the Baltics, the Czech
Republic, Hungary, and Poland quickly and robustly democratized after the
fall of communism. In Bulgaria, Romania, and Slovakia, in the second half

of the 1990s, EU leverage induced a turn toward liberalism and democracy
(Vachudova 2005, 2010). In the Western Hemisphere, pressures from the US and
the Organization of American States sometimes thwarted executive takeovers
and coups in the 1990s.
However, as Levitsky and Way (2010) have argued, international influences
are likely to have significant protective effect for democracy only in cases of high
vulnerability to external influence and strong linkages to the West—mainly,
countries in the Western Hemisphere and Eastern Europe. This includes only
one case in our volume—Argentina—although, as Riedl argues, international
pressures were crucial in the transition to democracy in South Africa.
What do the case studies in this volume tell us about the capacity of interna-
tional actors to sustain democracy in hard places? The only author who points
to international pressures and support as an important factor in sustaining what
he calls democratic moments is Lucan Way, writing on Georgia (2013–present),
Moldova (1991–2001), and Ukraine (1991–98, 2005–10). He asserts that a factor
in the emergence of democratic moments in the post-Soviet countries is that
anti-Russian leaders are susceptible to democratizing pressures. But, Way argues,
these international pressures have not been sufficient to nudge any of these three
countries into becoming even low-level electoral democracies. International
pressures were central to the decision of the apartheid regime elite to democ-
ratize in South Africa, but they are not a part of Riedl’s account of democratic
maintenance after the transition.
Mainwaring and Simison on Argentina and Bermeo on Timor-Leste explic-
itly note that international pressures and support are not a central part of their
explanations for democratic stability. Mainwaring and Simison argue that a
largely democratic neighborhood has supported democracy in Argentina, but
it is not the main part of their explanation for democratic survival. Bermeo
argues that democracy assistance was not a major contributing factor to dem-
ocratic endurance in Timor-Leste, although she writes that peacekeeping forces
were important. As many scholars have demonstrated, being part of a friendly
neighborhood increases a democracy’s chances of survival. However, except
for Argentina, the countries in our volume are not in neighborhoods that are
friendly to democracies. International constraints are not a major part of the
story of democratic longevity in these cases.
Beyond the case studies in this volume, in the new millennium, international
actors have been less successful in supporting democracy than they were in the
1990s. The US and the EU have been the main pro-democracy actors in the post–
Cold War period, but they have been hard-pressed to block incremental execu-
tive takeovers. The US and OAS were unable to prevent incremental democratic
erosions in Venezuela (1999–present), Bolivia (2006–19, followed by a coup in
late 2019), Ecuador (2007–17), and Nicaragua (2007–present), and they were
not able to thwart a coup in Honduras in 2009. The European Union has not
been able to prevent severe backsliding in Hungary since 2010 and Poland since
2015—countries that, based on their wealth, education levels, membership in
the EU, and proximity to Western Europe, should have been immune to erosion.
Chinese ascendance and Russia’s muscular support for authoritarian regimes has
further tipped the balance away from democracy. The ability of international ac-
tors to sustain democracies is more questionable now than it was in the 1990s.
Voters and Democratic Stability
A long lineage in some ways dating back to Tocqueville’s classic Democracy in

America has seen the broader political culture as responsible for maintaining
democracy. In recent decades, the work by Inglehart and Welzel (2005) exem-
plified the view that democratic voters make for democratic regimes (Almond
and Verba 1963; Inglehart 1997, 160–215). Nevertheless, as Claassen (2020)
noted, until his work, the empirical evidence to support these claims was thin.
Moreover, work that claimed that mass values accounted for democracy prior to
Claassen’s was subject to trenchant criticisms (Seligson 2002).
None of the authors in this volume argues that voters constrain politicians
to maintain democracy and thereby are responsible for sustaining democratic
regimes in hard places. This issue is the subject of an ongoing debate in the social
sciences, and it would be folly to attempt to resolve it here. Nevertheless, I would
be remiss if I did not say something about the subject because voters’ preferences
on some issues can hold politicians somewhat accountable.
In an excellent recent analysis, Claassen (2020) argued that public support
helps democracies survive. I have no doubt that strong democratic commitments
in the mass public help protect democracy and that, conversely, the absence of
such commitments makes democracy more vulnerable. Nevertheless, I am skep-
tical that voters’ attitudes are responsible for maintaining democratic regimes
in our cases or more generally. Several streams of literature underpin this as-
sumption. First, electoral accountability is blunt. Politicians in democracies do
not faithfully reflect mass preferences. Even when politicians claim to represent
mass publics, there are serious monitoring problems (Achen and Bartels 2016;
Przeworski et al. 1999). Politicians have significant autonomy and preferences
of their own, and elections do not suffice to induce them to mirror mass
preferences. As was displayed when President Trump claimed that widespread
fraud had cheated him out of the 2020 US presidential election and 70 percent of
Republican voters believed this assertion, politicians’ preferences and discourse
shape voter preferences and beliefs (Achen and Bartels 2016; Lenz 2012).
Even if voters were able to constrain politicians within tighter boundaries

than is feasible, it is not evident how much they would be willing to sacrifice
in other outcomes in order to achieve or preserve democracy. Recent work
by Graham and Svolik (2020) and Svolik (2019) based on survey experiments
raises serious doubts that US voters impose much of a penalty on politicians
for being flagrantly undemocratic. Their evidence suggests that the great
majority of voters care more about other things, especially in the context of
high polarization. At best, the evidence that voters care deeply about democ-
racy and push politicians and parties to sustain democracy is mixed. The
many executive takeovers led by authoritarian populists who repeatedly won
elections further suggests that large numbers of voters do not greatly value
democracy.
Along related lines, Bermeo (2003) argued that voters’ attitudes were not re-
sponsible for democratic breakdowns in Latin America and Europe; voters did
not defect from democracy, but key politicians and elite groups did. If this argu-
ment is correct, it suggests that even fairly widespread democratic commitments
among voters might not do much to protect democracy. Elites matter most for
democratic stability.
Because most voters in most democracies in hard places have less political
information and more pressing material needs than most voters in the advanced
industrial democracies, it seems even less likely that voters will highly priori-
tize democracy over substantive outcomes, and less likely that they will have the
means to push politicians into observing democratic norms.
Conclusion: External Constraints and Actors’ Motivations
This volume advances knowledge about what enables democracy to survive

in inhospitable terrain. I close with a few summary thoughts about how par-
tisan oppositions, institutional constraints, and actors’ own preferences sustain
democracy.
1. It is extremely helpful for democratic stability that the ruling party and the
other main parties be committed to democracy. Such parties exhibit what
Levitsky and Ziblatt (2018) call tolerance and mutual respect, and they
reject naked power grabs. If the governing parties are committed to de-
mocracy, the regime can survive for an extended time even in inhospitable
terrain.
2. Scholars don’t know a great deal about what generates and sustains nor-
mative preferences for democracy. This volume suggests that a widespread
repudiation of the terrible past and the memorialization of that terrible
past can help generate commitments to democracy that endure for an ex-
tended time. The chapters in this volume by Bermeo on Timor-Leste and
Mainwaring and Simison on Argentina each point to extremely costly
antecedent authoritarian regimes as generators of normative preferences
for democracy. The South African apartheid regime and British coloni-
alism in India generated similar widespread repudiation. The Argentine
case suggests that large-scale and ongoing efforts at memorializing human
rights abuses under dictatorship through human rights trials, public ed-
ucation, human rights museums, and human rights movements can help
nurture commitments to democracy even if government performance is
mediocre or worse.
3. Unfortunately, scholars and democracy advocates cannot always count
on the long-term sustainability of normative commitments to democracy.
In this volume, this is especially clear in Varshney’s chapter on India. The
solid commitment of the Congress Party to democracy during most of its
history—excepting the 1975–80 interlude—did not insulate democracy
from subsequent erosion when the BJP came to power.
4. Despite the large and illuminating outpouring of works on democratic ero-
sion and/or populism, we do not know a great deal about how to limit sup-
port for illiberal politicians and parties.
5. If the governing parties are naked power maximizers, democracy is not
likely to endure in countries with the challenges we describe in this volume.
Institutional constraints are not likely to be sufficient to hold democ-
racy together. Indeed, as scholars and democrats have discovered to great
alarm, even in developed countries with solid institutions, democracy is
not always able to withstand attacks from illiberal power maximizers. If all
politicians always pushed to constantly gain a political advantage (as some
do) at the expense of democracy, democracy in hard places would be an ex-
tremely unlikely equilibrium.
6. The volume opens, rather than resolves, the question of how best to sus-
tain normative commitments to democracy. Effective democratic govern-
ance is sometimes part of the answer, but in the new media environment
in which bizarre and far-fetched conspiracy theories can gain widespread
traction, it might not be enough.
7. A partisan balance of power is useful for sustaining democracy, but in
contexts of weak institutions, it is far from sufficient. Moreover, and per-
haps surprisingly, our volume suggests that a partisan balance of power
is far from necessary to sustain democracy even in difficult contexts. In
six countries analyzed in detail in this volume, democracy lasted for an
extended time even though the governing party or coalition sometimes
enjoyed indisputable electoral hegemony.
8. Strong legislatures and solid courts are helpful for sustaining democracy,
but they do not exist in most third-wave democracies—and certainly not
in most democracies in “hard places.” Strong legislatures are often not
effective as checks and balances if the executive’s party or coalition holds
a majority in the assembly.24 Moreover, in the medium term, they are
not likely to emerge in the third-wave democracies that do not already
have them.
9. The combination of a majority partisan opposition and strong
legislatures and courts probably makes democracy impregnable from
executive takeovers. However, because of the institutional weak-
ness of most legislatures and most judicial systems, this combination
does not exist in any of the countries studied in depth in this volume.
Surprisingly, democracy has survived for an extended time in six of
these countries despite often having the combination that seems
least propitious: weak institutions and frequent government majori-
ties. In my view, this highlights the importance of actors’ normative
commitments.
10. Institutional constraints must complement actors’ normative dispositions,
and vice versa. In all but well- established democracies, institutional
constraints can be vulnerable to rulers with authoritarian predilections. On
the flip side, democracy would be vulnerable if its existence always hinged
on the normative preferences of rulers and other actors.
Notes
* Many stimulating conversations and exchanges with Tarek Masoud and Aníbal
Pérez-Liñán shaped my thinking about this chapter. I am grateful to María Victoria
De Negri, Hannah Early Bagdanov, and Luis Elizondo Gracia for research assistance;
to Hannah Early Bagdanov, Nancy Bermeo, Benjamín García Holgado, Frances
Hagopian, Tarek Masoud, Ashutosh Varshney, and Kurt Weyland for comments; to
Jacob Nyrup and Stuart Bramwell for data on cabinet composition; and to Lucan
Way for information on governing parties in Georgia, Moldova, and Ukraine.
1. My understanding of actors includes organizations such as political parties, mili-
taries, labor and business associations, interest groups, and social movement organi-
zations, as well as heads of government and the ruling parties.
2. Some of the very distinguished authors who write in this volume offer divergent views.
3. The exact quote is “If men were angels, no government would be necessary. If angels
were to govern men, neither external nor internal controls on government would be
necessary” (Madison 1788).
4. Georgia, Moldova, and Ukraine are not cases of long-lasting democracy, as Lucan
Way’s chapter makes clear.
5. “Mitt Romney Blasted Trump’s Handling of the Pandemic and His Repeated Attacks
on Mail-In Voting,” New York Times, August 15, 2020, https://www.nytimes.com/
2020/08/15/world/coronavirus-covid-19.html#link-32696171.
6. Some exceptions deviate from this rule. In the US, the dominant faction of the
Southern Democrats rejected liberal democracy from before the Civil War
until around the 1960s. With the effort to overturn the results of the 2020 elec-
tion, Donald Trump and most of the contemporary Republican Party are also
exceptions.
7. Harvard political theorist Danielle Allen (2020, 60–61) wrote that “[p]‌eople can have
the chance of self-government through the institutions of constitutional democracy
if and only if they prioritize the preservation of these institutions over wins in sub-
stantive domains of policy.” My argument is not as demanding, but I agree that de-
mocracy is likely to be imperiled if the main political parties consistently prioritize
“wins in substantive domains of policy” above “the preservation of the institutions of
constitutional democracy.”
8. The National Party left the coalition government in 1995 because of its limited influ-
ence on policy-making.
9. The African National Congress was ideologically very heterogeneous. It included
some Leninists and some African nationalists who were indifferent to democracy.
10. In the dataset as a whole, 0.50 represents the median score. However, for democracy
to survive for an extended time in difficult circumstances, it is problematic if the gov-
erning party is only moderately committed to democratic norms. This is why I code
the median point on the scale as high illiberalism.
11. In contrast to Lührmann et al.’s (2020) coding for Georgia for the period after 2004,
Lucan Way (Chapter 5 of this volume) views the subsequent Saakashvili govern-
ment as competitive authoritarian. V-Party codes the victorious ENM party (United
National Movement) as 0.31 on the illiberalism scale in the 2004 elections.
12. An important body of literature has argued that high polarization erodes democratic
commitments. See Graham and Svolik 2020; Levitsky and Ziblatt 2018; McCoy et al.
2018; and Svolik 2019. Although the evidence for this argument is solid based on the
cases these authors analyze, among our nine country cases the relationship between
party system polarization and illiberalism as measured by V-Party data is weak. For
example, Indonesia combines mostly high illiberalism scores with consistently low
party system polarization. I measured party system polarization based on the V-Party
question about parties’ left-right position on a 0-to-6 scale (V2pariglef); each party’s
distance from the country’s mean weighted ideological score is weighted by its vote
share. At the country-election level for our nine countries, the bivariate correlation
between party system polarization and a weighted (again by vote share) illiberalism
score is only 0.01 (N =57 country elections). Moreover, contrary to Mainwaring and
Pérez-Liñán (2013), who showed a strong relationship between actors’ moderation
and commitment to democracy, the correlation between actors’ distance from the
ideological center and their illiberalism score is also very weak. At the party-election
year for our nine countries (n =265), the bivariate correlation between a party’s
ideological distance as an absolute value from the median distance of 3.00 and its
illiberalism score is only 0.15. In other words, party illiberalism is weakly related to
party distance from the center, and party system illiberalism is extremely weakly cor-
related with party system polarization.
13. Many politicians have willingly stepped down from office even when they enjoyed
high approval ratings. For example, every two-term US president until Franklin
D. Roosevelt willingly decided against running for a third term. Chilean President
Patricio Aylwin (1990–94) shortened his term, ruled out running for reelection, and
left office with extraordinarily high approval ratings. Many ruling parties have en-
gaged in what Levitsky and Ziblatt (2018) call forbearance; they do not consider using
every possible tool to expand their advantage. In the same book, Przeworski (2019,
20) talks about democratic and anti-democratic forces. If all politicians were willing
to do “whatever they can . . . to enhance their electoral advantage,” there is no differ-
ence between democratic and anti-democratic forces; there are no democrats.
14. Rustow (1970) claimed that democratic transitions often result from stalemates
among clashing actors under authoritarian rule. As a way of breaking the impasse,
the actors agree to transition to democracy even though none of them really wants
democracy. Although his article focused on transitions, and although Rustow himself
argued that democratic survival is a different issue, his logic could in principle apply
to survival as well.
15. On one of the few occasions in which the government proposed a measure that could
have weakened democracy, namely, a September 1988 draft bill to restrict freedom of
the press, the Congress Party rebelled; it was not willing to curb democratic rights.
The parliament walked out (Guha 2017, 593). It was not a partisan balance of power
but rather the Congress Party’s unwillingness to support a bill that would have weak-
ened freedom of the press that sank the measure.
16. Rustow was aware of this fact. He argued that democratic transitions could come
about without any actor being committed to democracy, but that democratic stabili-
zation rested on other processes.
17. The northern, Muslim-majority state of Jammu and Kashmir has been a notable
failure in this effort to combine a strong central government with substantial au-
tonomy at the state level. Kashmir has experienced frequent violent conflict and
frequent violations of citizen rights (Guha 2017, 237–57, 344–58, 609–12, 642–45;
Stepan et al. 2011, 109–15). In August 2019, the Modi government abrogated Jammu
and Kashmir’s autonomous status, which had been enshrined in Article 370 of the
Indian Constitution. De facto, this state now has a subnational authoritarian regime.
18. This is not to assert that federalism is consistently good for democracy. In the US,
for generations, southern states used claims of state rights to preserve authoritarian
enclaves and oppress African Americans (Mickey 2015). In Argentina, provincial po-
litical bosses have used a fiscal federal system with perverse incentives to build and
preserve provincial authoritarianism, which in turn has dragged down the country’s
level of democracy (Gervasoni 2018). One province of Indonesia, Aceh, has used its
special federal status as a way of defending sharia rule and downplaying the rights of
religious minorities.
19. Legislatures sometimes help impede executive takeovers; Gamboa (2017) analyzes
the example of Colombia under Álvaro Uribe (2002–10). In Argentina, the congress
blocked attempts to expand executive power by populist presidents Néstor Kirchner

(2003–7) and Cristina Fernández de Kircher (2007–15).
20. Reenock et al. (2013) found that independent courts favor democratic survival only
in contexts of moderately high economic development. Because most of our coun-
tries are poor, by implication, independent courts would not be expected to have an
impact on democratic survival. In some other studies, independent courts have had a
more consistently positive impact on democratic survival. See Gibler and Randazzo
(2011) and Pérez-Liñán and Mainwaring (2013).
21. 1 =Usually; 2 =About half the time; 3 =Seldom.
22. 1 =Seldom; 2 =About half the time; 3 =Usually.
23. Some scholars have reported that democracies are more likely to break down during
times of economic crisis (Gasiorowski 1995; Kapstein and Converse 2008, 58–68;
Przeworski et al. 2000; Slater et al. 2014, 362). Others have reported no effect (Alemán
and Yang 2011, 1140; Haggard and Kaufman 2016, 238–43; Mainwaring and Pérez-
Liñán 2013) or even that low growth was associated with a lower probability of break-
down (Remmer 1996). Still others have reported results that vary across different sets
of cases (Bernhard et al. 2001; Houle 2009, 612; Svolik 2015, 726).
24. As Wu (2020) wrote, “Structural checks can be overrated. The survival of our Republic
depends as much, if not more, on the virtue of those in government. . . . We have
grown too jaded about things like professionalism and institutions, and the idea of
men and women who take their duties seriously. . . . No external constraint can fully
substitute for the personal compulsion to do what is right.”
References
Abal Medina, Juan, and Julieta Súarez Cao. 2003. “Más allá del bipartidismo: El sistema
argentino de partidos.” Iberoamericana 3, no. 9 (March): 65–87.
Aboy Carlés, Gerardo. 2004. “Parque Norte o la doble ruptura alfonsinista.” In Historia
reciente: Argentina en democracia, edited by Marcos Novaro and Vicente Palermo, 11–
33. Buenos Aires: Edhasa.
Acemoglu, Daron, Simon Johnson, James A. Robinson, and Pierre Yared. 2008. “Income
and Democracy.” American Economic Review 98, no. 3: 808–842.
Acemoglu, Daron, and James A. Robinson. 2006. Economic Origins of Dictatorship and
Democracy. New York: Cambridge University Press.
Achen, Christopher H., and Larry M. Bartels. 2016. Democracy for Realists: Why Elections
Do Not Produce Responsive Government. Princeton, NJ: Princeton University Press.
Acuña, Carlos Hugo. 1995. “Intereses empresarios, dictadura y democracia en la
Argentina actual (o sobre por qué la burguesía abandona estrategias autoritarias y opta
por la estabilidad democrática).” In La nueva matriz política argentina, edited by Carlos
Hugo Acuña, 231–84. Buenos Aires: Ediciones Nueva Visión.
Acuña, Carlos Hugo. 2004. “The Industrial Bourgeoisie as a Political Actor: The Logic
of Its Organization and Strategies in Argentina (from Its XIX Century Origins to the
Present).” PhD diss., University of Chicago, Department of Political Science.
Acuña, Marcelo Luis. 1984. De Frondizi a Alfonsín: la tradición política del radicalismo. 2
vols. Buenos Aires: Centro Editor de América Latina.
Adejumobi, Said. 2000. “Elections in Africa: A Fading Shadow of Democracy?”
International Political Science Review 21, no. 1: 59–73.
Adjaho, Richard. 1992. La faillite du contrôle des finances publiques au Bénin (1960–1990).
Cotonou: Editions du Flamboyant.
Africa Confidential. 2002. “The Old Order Please.” Africa Confidential 43, no. 24
(December 6): 7.
Africa Confidential. 2006. “Twilight of the Chameleon.” Africa Confidential 47, no. 7
(March 31): 7.
Afrobarometer Data (South Africa and Benin, round 7, 2016–2018; accessed July 2021).
http://www.afrobarometer.org.
Agnes, Flavia. 2016. “Personal Laws.” In The Oxford Handbook of the Indian Constitution,
edited by Sujit Choudhry, Madhav Khosla, and Pratap Bhanu Mehta, 903– 20.
Oxford: Oxford University Press.
Ahmadov, Anar. 2014. “Oil, Democracy, and Context: A Meta-Analysis.” Comparative
Political Studies 47, no. 9: 1238–67.
Ahmed, Sadiq, and Ashutosh Varshney. 2012. “Battles Half Won: The Political Economy
of India’s Growth and Economic Policy.” In Oxford Handbook of the Indian Economy,
edited by Chetan Ghate, 56–102. Oxford: Oxford University Press.
Aivazaian, Anna. 2019. “ПЕРСПЕКТИВЫ ПАРЛАМЕНТСКИХ ВЫБОРОВ В
ГРУЗИИ В 2020 Г.” Научно-аналитический вестник ИЕ РАН 6, no. 12: 111–15.
270 References
Alemán, José, and David D. Yang. 2011. “A Duration Analysis of Democratic

Transitions and Authoritarian Backslides.” Comparative Political Studies 44, no. 9
(September): 1123–51.
Allen, Chris. 1992. “Restructuring an Authoritarian State: Democratic Renewal in Benin.”
Review of African Political Economy 19, no. 54: 42–58.
Allen, Danielle. 2020. “The Constitution Counted My Great- Great-
Grandfather as
Three-Fifths of a Free Person: Here’s Why I Love It Anyway.” The Atlantic 326, no. 3
(October): 58–63.
Almond, Gabriel A., and Sydney Verba. 1963. The Civic Culture: Political Attitudes and
Democracy in Five Nations. Princeton, NJ: Princeton University Press.
Amadeo, Eduardo. 1984. “Empresarios y democracia.” In Democracia, orden político
y parlamento, edited by Hilda Sabato and Marcelo Cavarozzi, 94– 103. Buenos
Aires: Centro Editor de América Latina.
Amuwo, ‘Kunle. 2003. “The State and the Politics of Democratic Consolidation in Benin,
1990–1999.” In Political Liberalization and Democratization in Africa, edited by Julius
Omozuanvbo Ihonvbere and John Mukum Mbaku, 141–78. Westport, CT: Praeger.
Anderson, Benedict. 1983. “Old State, New Society: Indonesia’s New Order in
Comparative Historical Perspective.” Journal of Asian Studies 42, no. 3 (May): 477–96.
Anderson, Benedict. 2001. “Imagining East Timor.” Lusotopie 8: 233–39.
Anderson, Benedict, Alief Djati, and Douglas Kammen. 2003. “Interview with Mário
Carrascalão.” Indonesia 76: 1–22.
Anderson, John. 1999. Kyrgyzstan: Central Asia’s Island of Democracy? Amsterdam,
ND: Harwood Academic Publishers.
Ansell, Ben, and David Samuels. 2010. “Inequality and Democratization: A Contractarian
Approach.” Comparative Political Studies 43, no. 12 (December): 1543–74.
Ansell, Ben, and David J. Samuels. 2014. Inequality and Democratization: An Elite-
Competition Approach. New York: Cambridge University Press.
Aprasidze, David. 2013. Nations in Transit 2012: Georgia. New York: Freedom House.
Arnold, Matthew B. 2009. “Challenges Too Strong for the Nascent State of Timor-
Leste: Petitioners and Mutineers.” Asian Survey 49, no. 3: 429–49.
Arthur, Catherine. 2019. Political Symbols and National Identity in Timor-Leste. Cham,
Switzerland: Palgrave Macmillan.
Aslaksen, Silje. 2010. “Oil and Democracy: More than a Cross-Country Correlation?”
Journal of Peace Research 47, no. 4: 421–31.
Aspinall, Edward, and Mark Berger. 2001. “The Break-Up of Indonesia? Nationalisms
after Decolonisation and the Limits of the Nation-State in Post-Cold War Southeast
Asia.” Third World Quarterly 22, no. 6 (December): 1003–24.
Aspinall, Edward, Allen Hicken, James Scambary, and Meredith Weiss. 2018.
“Exchange: Timor-Leste Votes—Parties and Patronage.” Journal of Democracy 29, no.
1 (January): 153–67.
Austin, Granville. 1966. The Indian Constitution: Cornerstone of a Nation. Delhi: Oxford
University Press.
Auyero, Javier. 2007. Routine Politics and Violence in Argentina: The Gray Zone of State
Power. New York: Cambridge University Press.
Balcells, Laia, Valeria Palanza, and Elsa Voytas. 2022. “Do Transitional Justice Museums
Persuade Visitors? Evidence from a Field Experiment.” Journal of Politics 84, no. 1
(January): 496–510. https://doi.org/10.1086/714765.
References 271
Ballard, John. 2008. Triumph of Self-Determination: Operation Stabilise and United

Nations Peace-Making in East Timor. Westport, CT: Greenwood Publishing Group.
Ballester, Horacio. 1996. Memorias de un coronel democrático: medio siglo de historia
política argentina en la óptica de un militar. Buenos Aires: Ediciones de la Flor.
Banégas, Richard. 2003. La démocratie à pas de caméléon. Transition et imaginaires
politiques au Bénin. Paris: Karthala.
Banégas, Richard. 2014. “Briefing Benin: Challenges for Democracy.” African Affairs 113,
no. 452: 449–59.
Barkan, Joel D. 1995. “Debate: PR and Southern Africa: Elections in Agrarian Societies.”
Journal of Democracy 6, no. 4: 106–16.
Basu, Deepankar. 2021. “Majoritarian Politics and Hate Crimes against Religious
Minorities: Evidence from India, 2009–2018.” World Development 146: 105540. https://
doi.org/10.1016/j.worlddev.2021.105540.
Bates, Robert H. 1984. Markets and States in Tropical Africa: The Political Basis of
Agricultural Policies. Berkeley: University of California Press.
Battaglino, Jorge. 2010. “La política militar de Alfonsín: la implementación del control
civil en un contexto desfavorable.” In Discutir Alfonsín, edited by Roberto Gargarella,
María Victoria Murillo, and Mario Pecheny, 161–84. Buenos Aires: Siglo XXI.
Bayart, Jean-François. 1989. L’état en Afrique. La politique du ventre. Paris: Fayard.
Beissinger, Mark R. 2002. Nationalist Mobilization and the Collapse of the Soviet State.
New York: Cambridge University Press.
Bellin, Eva. 2004. “The Robustness of Authoritarianism in the Middle East: Exceptionalism
in Comparative Perspective.” Comparative Politics 36, no. 2: 139–57.
Bellin, Eva. 2012. “Reconsidering the Robustness of Authoritarianism in the Middle
East: Lessons from the Arab Spring.” Comparative Politics 44, no. 2: 127–49.
Beltrán, Gastón J. 2006. “Acción empresaria e ideología. La génesis de las reformas
estructurales.” In Los años de Alfonsín, ¿el poder de la democracia o la democracia del
poder?, edited by Alfredo Raúl Pucciarelli, 199–243. Buenos Aires: Siglo XXI.
Berbecel, Dan. 2022. Presidential Power in Latin America: Examining the Cases of
Argentina and Chile. New York: Routledge.
Berman, Sheri. 1998. The Social Democratic Movement: Ideas and Politics in the Making of
Interwar Europe. Cambridge, MA: Harvard University Press.
Bermeo, Nancy. 1992. “Democracy and the Lessons of Dictatorship.” Comparative Politics
24, no. 3 (April): 273–91.
Bermeo, Nancy. 2003. Ordinary People in Extraordinary Times. Princeton, NJ: Princeton
University Press.
Bermeo, Nancy. 2010. “Armed Conflict and the Durability of Electoral Democracy.” In
In War’s Wake: International Conflict and the Fate of Liberal Democracy, edited by
Elizabeth Kier and Ronald Krebs, 67–94. New York: Cambridge University Press.
Bermeo, Nancy. Forthcoming. Democracy after War. Princeton, NJ: Princeton
University Press.
Bermeo, Nancy, and Deborah J. Yashar, eds. 2016a. Parties, Movements, and Democracy in
the Developing World. New York: Cambridge University Press.
Bermeo, Nancy, and Deborah Yashar. 2016b. “Parties, Movements, and the Making
of Democracy.” In Parties, Movements, and Democracy in the Developing World,
edited by Nancy Bermeo and Deborah Yashar, 1– 27. New York: Cambridge
University Press.
272 References
Bernhard, Michael, Timothy Nordstrom, and Christopher Reenock. 2001. “Economic

Performance, Institutional Intermediation and Democratic Breakdown.” Journal of
Politics 63 (3): 775–803.
Bertrand, Jacques. 2007. “Indonesia’s Quasi- Federalist Approach: Accommodation
among Strong Integrationist Tendencies.” International Journal of Constitutional Law
5, no. 4 (October): 576–605.
Bexley, Angie, and Nuno Rodrigues Tchailoro. 2013. “Consuming Youth: Timorese in the
Resistance against Indonesian Occupation.” Asia Pacific Journal of Anthropology 14, no.
5: 405–22.
Bierschenk, Thomas. 2009. “Democratization without Development: Benin 1989–2009.”
International Journal of Politics, Culture, and Society IJPS 22, no. 3: 337–57.
Bierschenk, Thomas, Elizabeth Thioléron, and Nassirou Bako-Arifari. 2003. “Benin.”
Development Policy Review 21, no. 2: 161–78.
Bockelie, Josette. 2013. “Democratic Consolidation in Benin: A Study of Socio-Economic
and Political Development.” Undergraduate thesis, Paris Institute of Political Studies.
Boix, Carles. 2003. Democracy and Redistribution. Cambridge: Cambridge University Press.
Boix, Carles. 2011. “Democracy, Development, and the International System.” American
Political Science Review 105, no. 4 (November): 809–28.
Boix, Carles, and Susan C. Stokes. 2003. “Endogenous Democratization.” World Politics
55, no. 4 (July): 517–49.
Boix-Miller-Rosato (BMR). dichotomous coding of democracy, 1800–2015 (Version 3.0).
Accessed August 2019. https://sites.google.com/site/mkmtwo/data.
Bolt, Jutta, Robert Inklaar, Herman de Jong, and Jan Luiten van Zanden. 2018. “Maddison
Project Database, version 2018.” Accessed July 2018. www.ggdc.net/maddison.
Bonvecchi, Alejandro. 2002. “Estrategia de supervivencia y tácticas de disuasión. Los
procesos políticos de la política económica después de las reformas estructurales.” In
El derrumbe político en el ocaso de la convertibilidad, edited by Marcos Novaro, 107–93.
Buenos Aires: Norma.
Booysen, Susan. 2014. Twenty Years of South African Democracy: Citizen Views of Human
Rights, Governance and the Political System. Washington, DC: Freedom House. https://
freed omho use.org/ s ites/ d efa u lt/ f iles/ Twe nty%20Ye ars%20of%20So uth%20Afri
can%20Democracy.pdf.
Booysen, Susan, and Grant Masterson. 2009. “South Africa.” In Compendium of Elections
in Southern Africa 1989–2009: 20 Years of Multiparty Democracy, edited by Denis
Kadima and Susan Booysen, 399–402. Johannesburg: EISA.
Bosoer, Fabián. 2006. “¿Crisis con salvataje? Una hipótesis sobre el componente
parlamentario en el auto-rescate de las democracias sudamericanas: 1999–2004.” In
Instituciones, democracia e integración regional en el Mercosur, edited by Giorgio Alberti,
Elsa Llenderrozas, and Julio Pinto, 245–66. Buenos Aires: Bononiae Libris-Prometeo.
Botana, Natalio. 2006. Poder y hegemonía. el régimen político después de la crisis. Buenos
Aires: Emecé.
Bovensiepen, Judith, and Maj Nygaard-Christensen. 2018. “Petroleum Planning as State
Building in Timor-Leste.” Asia Pacific Journal of Anthropology 19, no. 5: 412–31.
Boyle, Michael. 2014. Violence after War: Explaining Instability in Post-Conflict States.
Baltimore, MD: JHU Press.
Braithwaite, John, Hilary Charlesworth, and Adérito Soares. 2012. Networked Governance
of Freedom and Tyranny: Peace in Timor-Leste. Canberra, Australia: ANU Press.
References 273
Bratton, Michael, and Chris Landsberg. 1998. “Aiding Reconstruction and Development
in South Africa: Promise and Delivery.” In Good Intentions: Pledges of Aid for Postconflict
Recovery, edited by Shepard Forman and Stewart Patrick, 259–314. Boulder, CO: Lynne
Rienner Publishers.
Bratton, Michael, and Nicholas van de Walle. 1997. Democratic Experiments in
Africa: Regime Transitions in Comparative Perspective. Cambridge: Cambridge
University Press.
Brinks, Daniel, and Michael Coppedge. 2006. “Diffusion Is No Illusion: Neighbor
Emulation in the Third Wave of Democracy.” Comparative Political Studies 39, no. 4
(May): 463–89.
Brinks, Daniel M., Steven Levitsky, and María Victoria Murillo. 2019. Understanding Institutional
Weakness: Power and Design in Latin American Institutions. New York: Cambridge
University Press.
Brownlee, Jason. 2007. Durable Authoritarianism in an Age of Democratization. New York:
Cambridge University Press.
Brownlee, Jason, Tarek Masoud, and Andrew Reynolds. 2013. “Tracking the ‘Arab
Spring’: Why the Modest Harvest?” Journal of Democracy 24, no. 4: 29–44.
Bruno, Ángel. 1984. “Derechos humanos y el parlamento.” In Democracia, orden
político y parlamento, edited by Hilda Sabato and Marcelo Cavarozzi, 127–35. Buenos
Aires: Centro Editor de América Latina.
Budiardjo, Carmel, and Soei Liong Liem. 1984. The War against East Timor.
London: Zed Books.
Bukia, Sopho. 2015. “Georgian TV Dispute Gets Nasty.” Institute for War and Peace
Reporting, Congressional Research Service, Issue 794 (October 27). https://www.
refworld.org/docid/564c41a64.html.
Bureau of European and Eurasian Affairs. 2009. “Foreign Operations Appropriated
Assistance: Belarus.” Washington, DC: US Department of State.
Burnell, Peter J. 2010. Is There a New Autocracy Promotion? Madrid: Fride.
Capizzi, Elaine, Helen Hill, and Dave Macey. 1976. “FRETILIN and the Struggle for
Independence in East Timor.” Race & Class 17, no. 4: 381–95.
Caplan, Richard. 2005. International Governance of War-Torn Territories. New York:
Oxford University Press.
Carey, John M., and Matthew Soberg Shugart, eds. 1998. Executive Decree Authority.
Carothers, Thomas. 1999. Aiding Democracy Abroad: The Learning Curve. Washington
DC: Carnegie Endowment.
Cavarozzi, Marcelo. 1983. Autoritarismo y democracia (1955–1983). Buenos Aires: Centro
Editor de América Latina.
CAVR (Timor-Leste Commission for Reception, Truth, and Reconciliation). 2005. Chega!
The Report of the Commission for Reception, Truth, and Reconciliation in Timor‐Leste.
Part 3. http://www.etan.org/etanpdf/2006/CAVR/03-History-of-the-Conflict.pdf.
CBM (Consultative Business Movement). 1997. Submission to the Truth and Reconciliation
Commission of South Africa. Johannesburg, South Africa: CBM.
Cecire, Michael. 2013. “Georgia’s 2012 Elections and Lessons for Democracy Promotion.”
Orbis 57, no. 2 (Spring): 232–50.
Cederman, Lars-Erik, Andreas Wimmer, and Brian Min. 2010. “Why Do Ethnic Groups
Rebel? New Data and Analysis.” World Politics 62, no. 1 (January): 87–119.
274 References
CEPALSTAT. Databases and Statistical Publications. Accessed January 2019. http://estad

isticas.cepal.org/cepalstat.
Chabi, Maurice. 1993. Banqueroute: Mode d’emploi (Un marabout dans les griffes de la
mafia béninoise). Cotonou: Editions Gazette Livres.
Chakrabarti, Poulomi. 2019. “One Nation, Many Worlds: Education, Health and Dignity
in India.” PhD diss., Brown University.
Challenor, Herschelle. 1970. “French-Speaking West Africa’s Dahomeyan Strangers in
Colonization and Decolonization.” PhD diss., Columbia University.
Charosky, Hernán. 2002. “Honestos y audaces: realizaciones y límites de la política
anticorrupción.” In El derrumbe político en el ocaso de la convertibilidad, edited by
Marcos Novaro, 195–252. Buenos Aires: Norma.
Chauchard, Simon. 2017. Why Representation Matters: Ethnic Quotas in Rural India.
Cheeseman, Nic. 2015. Democracy in Africa: Successes, Failures, and the Struggle for
Political Reform. New York: Cambridge University Press.
Cheresky, Isidoro. 2008. Poder presidencial, opinión pública y exclusión social. Buenos
Aires: Ediciones Manantial.
Chernykh, Svitlana, David Doyle, and Timothy J. Power. 2017. “Measuring Legislative
Power: An Expert Reweighting of the Fish-Kroenig Parliamentary Powers Index.”
Legislative Studies Quarterly 42, no. 2 (May): 295–320.
Chhibber, Pradeep. 2014. Religious Practice and Democracy in India. New York: Cambridge
University Press.
Chopra, Jarat. 2000. “The UN’s Kingdom in East Timor.” Survival 42, no. 3 (Autumn): 27–29.
Chopra, Jarat. 2002. “Building State Failure in East Timor.” Development and Change 33,
no. 5: 979–1000.
CIA (Central Intelligence Agency). 2021. “Georgia.” Last updated September 22, 2021.
Accessed October 10, 2021. https://www.cia.gov/the-world-factbook/countries/
georgia/.
CID (Consultoría Interdisciplinaria en Desarrollo S. A.). 1985. USIA Poll # 1985-
I85110: Youth Attitudes Survey. Buenos Aires: Consultoría Interdisciplinaria en
Desarrollo S. A.
Claassen, Christopher. 2020. “Does Public Support Help Democracy Survive?” American
Journal of Political Science 64, no. 1 (January): 118– 34. https://doi.org/10.1111/
ajps.12452.
CONADEP (Comisión Nacional sobre la Desaparición de Personas). 1984. Nunca
más: informe final de la Comisión Nacional sobre la Desaparición de Personas. Buenos
Aires: EUDEBA.
Coppedge, Michael. 2002. “Theory Building and Hypothesis Testing: Large-vs. Small-N
Research on Democratization.” Paper prepared for presentation at the Annual Meeting
of the Midwest Political Science Association, Chicago, IL, April 25–27.
Coppedge, Michael, Benjamin Denison, Paul Friesen, Lucia Tiscornia, and Yang Xu.
2022. “International Influences: The Hidden Dimension.” In Why Democracies Develop
and Decline, edited by Michael Coppedge, Amanda Edgell, Carl Henrik Knutsen, and
Staffan I. Lindberg. New York: Cambridge University Press.
Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Svend-
Erik Skaaning, Jan Teorell, David Altman, Michael Bernhard, M. Steven Fish, Agnes
Cornell, Sirianne Dahlum, Haakon Gjerløw, Adam Glynn, Allen Hicken, Joshua
Krusell, Anna Lührmann, Kyle L. Marquardt, Kelly McMann, Valeriya Mechkova, Juraj
References 275
Medzihorsky, Moa Olin, Pamela Paxton, Daniel Pemstein, Josefine Pernes, Johannes
von Römer, Brigitte Seim, Rachel Sigman, Jeffrey Staton, Natalia Stepanova, Aksel
Sundström, Eitan Tzelgov, Yi-ting Wang, Tore Wig, Steven Wilson, and Daniel Ziblatt.
2018. “V-Dem Country-Year Dataset v8.” Varieties of Democracy (V-Dem) Project.
https://doi.org/10.23696/vdemcy18.
Coppedge, Michael, John Gerring, Carl Henrik Knutsen, Staffan I. Lindberg, Jan Teorell,
David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Anna
Luhrmann, Kyle L. Marquardt, Kelly McMann, Pamela Paxton, Daniel Pemstein,
Brigitte Seim, Rachel Sigman, Svend-Erik Skaaning, Jeffrey Staton, Agnes Cornell, Lisa
Gastaldi, Haakon Gjerløw, Valeriya Mechkova, Johannes von Römer, Aksel Sundtröm,
Eitan Tzelgov, Luca Uberti, Yi-ting Wang, Tore Wig, and Daniel Ziblatt. 2020a. “V-
Dem Codebook v10.” Varieties of Democracy (V-Dem) Project. https://www.v-dem.
net/media/filer_public/28/14/28140582-43d6-4940-948f-a2df84a31893/v-dem_c
odebook_v10.pdf.
David Altman, Michael Bernhard, M. Steven Fish, Adam Glynn, Allen Hicken, Anna
Luhrmann, Kyle L. Marquardt, Kelly McMann, Pamela Paxton, Daniel Pemstein,
Brigitte Seim, Rachel Sigman, Svend-Erik Skaaning, Jeffrey Staton, Steven Wilson,
Agnes Cornell, Nazifa Alizada, Lisa Gastaldi, Haakon Gjerløw, Garry Hindle, Nina
Ilchenko, Laura Maxwell, Valeriya Mechkova, Juraj Medzihorsky, Johannes von
Römer, Aksel Sundström, Eitan Tzelgov, Yi-ting Wang, Tore Wig, and Daniel Ziblatt.
2020b. “V-Dem Country–Year Dataset v10.” Varieties of Democracy (V-Dem) Project.
https://doi.org/10.23696/vdemds20.
Nazifa Alizada, David Altman, Michael Bernhard, Agnes Cornell, M. Steven Fish,
Lisa Gastaldi, Haakon Gjerløw, Adam Glynn, Allen Hicken, Garry Hindle, Nina
Ilchenko, Joshua Krusell, Anna Luhrmann, Seraphine F. Maerz, Kyle L. Marquardt,
Kelly McMann, Valeriya Mechkova, Juraj Medzihorsky, Pamela Paxton, Daniel
Pemstein, Joseﬁne Pernes, Johannes von Römer, Brigitte Seim, Rachel Sigman, Svend-
Erik Skaaning, Jeffrey Staton, Aksel Sundström, Ei-tan Tzelgov, Yi-ting Wang, Tore
Wig, Steven Wilson and Daniel Ziblatt. 2021. “V-Dem Country–Year Dataset v11.1.”
Varieties of Democracy (V-Dem) Project. https://doi.org/10.23696/vdemds21.
Cornell, Svante E. 2014. “Getting Georgia Right.” European View 13: 185.
Corrales, Javier. 2018. Fixing Democracy: Why Constitutional Change Often Fails to
Enhance Democracy in Latin America. New York: Oxford University Press.
Coveney, Joseph. 2021. “FIRTHLOGIT: Stata Module to Calculate Bias Reduction in
Logistic Regression.” https://econpapers.repec.org/software/bocbocode/s456948.htm.
Cox, Gary, and Scott Morgenstern. 2001. “Latin America’s Reactive Assemblies and
Proactive Presidents.” Comparative Politics 33, no. 2 (January): 171–89.
Crenzel, Emilio. 2015. “Genesis, Uses, and Significations of the Nunca Más Report in
Argentina.” Latin American Perspectives 42, no. 3: 20–38. https://doi.org/10.1177/
0094582X15570875.
Cristalis, Irena. 2009. East Timor: A Nation’s Bitter Dawn. London: Zed Books.
Croissant, Aurel. 2008. “The Perils and Promises of Democratization through United
Nations Transitional Authority— Lessons from Cambodia and East Timor.”
Democratisation 15, no. 3: 649–68.
Croissant, Aurel, and David Kuehn. 2009. “Patterns of Civilian Control of the Military in
East Asia’s New Democracies.” Journal of East Asian Studies 9, no. 2: 187–217.
276 References
Crowther, William. 1991. “The Politics of Ethno-National Mobilization: Nationalism and

Reform in Soviet Moldavia.” Russian Review 50, no. 2: 183–202.
Crowther, William. 2007. “Moldova, Transnistria, and the PCRM’s Turn to the West.” East
European Quarterly 61, no. 3: 273–304.
Dahl, Robert A. 1971. Polyarchy: Participation and Opposition. New Haven: Yale University Press.
Dahl, Robert A. 1973. Polyarchy: Participation and Opposition. New Haven: Yale
University Press.
Dahl, Robert A. 1989. Democracy and Its Critics. New Haven: Yale University Press.
Darwich, May. 2017. “Creating the Enemy, Constructing the Threat: The Diffusion of
Repression against the Muslim Brotherhood in the Middle East.” Democratization 24
(7): 1289–1306.
David, Pauline. 2020. “TBC Bank Case: Money Laundering Prosecution.” Transparency
International. https://transparency.ge/en/post/assessment-money-laundering-pros
ecution-mamuka-khazaradze-badri-japaridze-and-avtandil.
de Ipola, Emilio. 1987. “La difícil apuesta del peronismo democrático.” In Ensayos sobre la
transición democrática en la Argentina, edited by José Nun and Juan Carlos Portantiero,
333–74. Buenos Aires: Puntosur Editores.
de Jager, Nicola, and Cindy Steenekamp. 2019. “Political Radicalism: Responding to the
Legitimacy Gap in South Africa.” In Democracy under Threat: A Crisis of Legitimacy?,
edited by Ursula van Beek, 147–69. Cham, Switzerland: Palgrave Macmillan.
De Riz, Liliana. 1982. “Argentina: ni democracia estable ni régimen militar sólido.” Revista
Mexicana de Sociología 44, no. 4: 1203–23. https://doi.org/10.2307/3540128.
de Souza, Peter, Suhas Palshikar, and Yogendra Yadav, eds. 2008. The State of Democracy in
South Asia: A Report. Delhi: Oxford University Press.
Decalo, Samuel. 1976. Coups and Army Rule in Africa Motivations and Constraints. New
Haven: Yale University Press.
Decalo, Samuel. 1997. “Benin: The First of the New Democracies.” In Political Reform in
Francophone Africa, edited by John F. Clark and David E. Gardinier, 43–61. Boulder,
CO: Westview Press.
Decalo, Samuel. 2018. “Benin: First of the New Democracies.” In Political Reform
in Francophone Africa, edited by John Clark and David Gardinier, 43– 61.
New York: Routledge.
Devdariani, Jaba. 2003. “Georgia: Security at Stake.” Transitions Online, April 9.
Devdariani, Jaba. 2004. “Georgia: The Year of Revolution.” Transitions Online, April 15.
Diamint, Rut. 2006. “Crisis, Democracy, and the Military in Argentina.” In Broken
Promises? The Argentine Crisis and Argentine Democracy, edited by Edward Epstein
and David Pion-Berlin, 163–79. Lanham, MD: Lexington Books.
Diamond, Larry, Seymour Martin Lipset, and Juan Linz. 1987. “Building and Sustaining
Democratic Government in Developing Countries: Some Tentative Findings.” World
Affairs 150, no. 1 (Summer): 5–19.
Dinerstein, Ana Cecilia. 2004. “Más allá de la crisis. Acerca de la naturaleza del cambio
político en Argentina.” Revista Venezolana de Economía y Ciencias Sociales 10, no. 1
(January–April): 241–69.
Ding, Iza, and Dan Slater. 2021. “Democratic Decoupling.” Democratization 28, no. 1
(January): 63–80.
Diop, Omar. 2006. Partis politiques et processus de transition démocratique en Afrique
noire: Recherches sur les enjeux juridiques et sociologiques du multipartisme dans
quelques pays de l’espace francophone. Paris: Editions Publibook des Ecrivans.
References 277
Divan, Shyam. 2016. “Public Interest Litigation.” In The Oxford Handbook of the Indian
Constitution, edited by Sujit Choudhry, Madhav Khosla, and Pratap Bhanu Mehta,
662–79. Oxford: Oxford University Press.
Dolidze, Ana. 2007. “Inside Track: Georgia’s Path to Authoritarianism.” National Interest
Online, August 24.
Donno, Daniela, and Bruce Russett. 2004. “Islam, Authoritarianism, and Female
Empowerment: What Are the Linkages?” World Politics 56, no. 4 (July): 582–607.
Doyon, Louise. 1988. “El crecimiento sindical bajo el peronismo.” In La formación del
sindicalismo peronista, edited by Juan Carlos Torre, 169–81. Buenos Aires: Legasa.
Drazanova, Lenka. 2019. “Historical Index of Ethnic Fractionalization Dataset (HIEF).”
https://doi.org/10.7910/DVN/4JQRCL, Harvard Dataverse, V2, UNF:6:z4J/b/
PKbUpNdIoeEFPvaw== [fileUNF].
Dunn, James. 2003. East Timor: A Rough Passage to Independence. New South Wales:
Longueville Books.
Durand, Frédéric. 2016. History of Timor-Leste. Chiang Mai: Silkworm Books.
ECLAC (Economic Commission for Latin America and the Caribbean). 1985. “Nota
sobre la evolución de la economía argentina en 1984.” Versión preliminar. Documento
de trabajo nro. 15, ECLAC, Buenos Aires. https://repositorio.cepal.org/bitstream/han
dle/11362/28507/S8500616_es.pdf.
ECLAC (Economic Commission for Latin America and the Caribbean). 1993. Social
Panorama of Latin America 1993. Santiago, Chile: ECLAC.
ECLAC (Economic Commission for Latin America and the Caribbean). 2001. Panorama
social de América Latina 2000–2001. Santiago, Chile: Naciones Unidas.
Panorama of Latin America 2002–2003. Santiago, Chile: ECLAC.
Panorama of Latin America 2005. Santiago, Chile: ECLAC. https://www.cepal.org/en/
publications/1224-social-panorama-latin-america-2005.
Panorama of Latin America 2020. Santiago, Chile: ECLAC.
Emery, Alan. 2006. “Privatization, Neoliberal Development, and the Struggle for Workers’
Rights in Post-Apartheid South Africa.” Social Justice 33, no. 3 (September): 6–19.
Englebert, Pierre. 1996. “Benin: Recent History.” In Africa South of the Sahara 1996.
London: Europa Publications.
Englebert, Pierre. 2002. State Legitimacy and Development in Africa. Boulder, CO: Lynne
Rienner Publishers.
Epstein, David L., Robert Bates, Jack Goldstone, Ida Kristensen, and Sharyn O’Halloran. 2006.
“Democratic Transitions.” American Journal of Political Science 50, no. 3 (July): 551–69.
Fairbanks, Charles H., Jr. 2004. “Georgia’s Rose Revolution.” Journal of Democracy 15, no.
2: 110–24.
Fairbanks Charles H., Jr., and Alexi Gugushvili. 2013. “A New Chance for Georgian
Democracy.” Journal of Democracy 24, no. 1 (January): 116–27.
FBIS (Foreign Broadcast Information Service). 1992. “Daily Reports, Central Eurasia.”
FBIS-SOV-92-009, January 14, page 62. Naples, FL: Readex.
FBIS (Foreign Broadcast Information Service). 1994. “Daily Reports, Central Eurasia.”
August 3, page 38. Naples, FL: Readex.
Fealy, Greg. 2021. “Jokowi in the Covid-19 Era: Repressive Pluralism, Dynasticism, and the
Overbearing State.” Bulletin of Indonesian Economic Studies 56, no. 3 (September): 301–23.
278 References
Feijó, Rui Graça. 2016. Dynamics of Democracy in Timor-Leste: The Birth of a Democratic
Nation, 1999–2012. Amsterdam: Amsterdam University Press.
Feijó, Rui Graça. 2018. “Timor-Leste in 2017: Between a Diplomatic Victory and the
Return of ‘Belligerent Democracy.’” Asian Survey 58, no. 1: 206–12.
Fernandes, Clinton. 2011. The Independence of East Timor: Multi- dimensional
Perspectives—Occupation, Resistance, and International Political Activism. Portland,
OR: Sussex Academic Books.
Figueiredo, Argelina Cheibub, and Fernando Limongi. 1999. Executivo e Legislativo na
Nova Ordem Constitucional. Rio de Janeiro: Editora FGV.
Fish, M. Steven. 1995. Democracy from Scratch: Opposition and Regime in the New Russian
Revolution. Princeton, NJ: Princeton University Press.
Fish, M. Steven. 2002. “Islam and Authoritarianism.” World Politics 55, no. 1
(October): 4–37.
Fish, M. Steven, and Robin S. Brooks. 2004. “Does Diversity Hurt Democracy?” Journal of
Democracy 15, no. 1 (January): 154–66.
Franco, Marina. 2009. “‘Homeland Security’ as a State Policy in 1970’s Argentina.”
Antíteses 2, no. 4: 887–914.
Freedom House. 2002. Freedom of the Press: Argentina. New York: Freedom House.
Freedom House. 2013. Nations in Transit 2013—Georgia. New York: Freedom House.
Freedom House. 2019a. “Country and Territory Ratings and Statuses, 1973–2019.”
Accessed July 2019. https://freedomhouse.org/report/freedom-world.
Freedom House. 2019b. Nations in Transit 2019—Georgia. New York: Freedom House.
Freedom House. 2020a. “Freedom in the World 2020: Benin.” Accessed July 2021. https://
freedomhouse.org/country/benin/freedom-world/2020.
Freedom House. 2020b. Nations in Transit 2020—Georgia. New York: Freedom House.
Freedom House. 2021a. “Country and Territory Ratings and Statuses, 1973–2021.”
Accessed April 23, 2021. https://freedomhouse.org/sites/default/files/2021-02/
Country_and_Territory_Ratings_and_Statuses_FIW1973-2021.xlsx.
Freedom House. 2021b. Freedom in the World 2021: Democracy under Siege. Washington,
DC: Freedom House.
Freedom House. n.d. “Global Freedom Scores.” Countries and Territories. Accessed
October 10, 2021. https://freedomhouse.org/countries/freedom-world/scores?sort=
desc&order=Total%20Score%20and%20Status.
Frey, Bruno S., Matthias Benz, and Alois Stutzer. 2004. “Introducing Procedural Utility: Not
Only What, but Also How Matters.” Journal of Institutional and Theoretical Economics
(JITE) 160, no. 3 (September): 377–401. https://doi.org/10.1628/0932456041960560.
Friedrich, Carl Joachim. 1939. “Democracy and Dissent.” The Political Quarterly 10, no. 4
(October): 571–82.
Fuller, Liz. 2005. “Some Georgian Journalists Feel Less Equal than Others.” RFE/RL
Report 5, no. 1 (January 3).
Fumagalli, Matteo. 2014. “The 2013 Presidential Election in the Republic of Georgia.”
Electoral Studies 35 (September): 395–97.
Galston, William. 2018. “The Populist Challenge to Liberal Democracy.” Journal of
Democracy 29, no. 2 (April): 5–19.
References 279
Gama, Paulino. 1995. “The War in the Hills, 1975– 85: A Fretilin Commander
Remembers.” In East Timor at the Crossroads: The Forging of a Nation, edited by Peter
Carey and Carter Bentley, 97–105. Honolulu: University of Hawaii Press.
Gamboa, Laura. 2017. “Opposition at the Margins: Strategies against the Erosion of
Democracy in Colombia and Venezuela.” Comparative Politics 49, no. 4 (July): 457–77.
Garay, Candelaria. 2016. Social Policy Expansion in Latin America. New York: Cambridge
University Press.
García Lerena, Roberto. 2007. Saúl Ubaldini: Crónicas de un militante obrero peronista.
Buenos Aires: Ediciones Runa Comunicaciones.
García Vázquez, Enrique. 1994. “La economía durante la presidencia de Illia.” Desarrollo
Económico 34, no. 134: 291–95. https://doi.org/10.2307/3467321.
Garrard, John. 2002. Democratisation in Britain: Elites, Civil Society and Reform since
1800. Basingstoke, Hampshire: Palgrave.
Gasiorowski, Mark J. 1995. “Economic Crisis and Political Regime Change: An Event
History Analysis.” American Political Science Review 89, no. 4 (December): 882–97.
Gazibo, Mamoudou. 2005. “Foreign Aid and Democratization: Benin and Niger
Compared.” African Studies Review 48, no. 3: 67–87.
Gazibo, Mamoudou. 2013. “Beyond Electoral Democracy: Foreign Aid and the Challenge
of Deepening Democracy in Benin.” In Democratic Trajectories in Africa: Unravelling
the Impact of Foreign Aid, edited by Danielle Resnick and Nicolas van de Walle, 228–55.
Geddes, Barbara. 1999. “What Do We Know about Democratization after Twenty Years?”
Annual Review of Political Science 2: 115–44.
Geddes, Barbara, Joseph Wright, and Erica Frantz. 2014. “Autocratic Breakdown and
Regime Transitions: A New Data Set.” Perspectives on Politics 12, no. 2: 313–31.
Gente, Regis. 2013. “Bidzina Ivanishvili, a Man Who Plays According to Russian Rules?”
Caucasus Survey 1, no. 1 (October): 1–9.
Gerchunoff, Pablo, and Lucas Llach. 2010. El ciclo de la ilusión y el desencanto. Un siglo de
políticas económicas argentinas. Buenos Aires: Emecé.
Gervasoni, Carlos. 2018. Hybrid Regimes within Democracies Fiscal Federalism and
Subnational Rentier States. New York: Cambridge University Press.
Gibler, Douglas M., and Kirk A. Randazzo. 2011. “Testing the Effects of Independent
Judiciaries on the Likelihood of Democratic Backsliding.” American Journal of Political
Science 55, no. 3 (July): 696–709.
Gibson, Edward L. 1996. Class and Conservative Parties: Argentina in Comparative
Perspective. Reprint edition. Baltimore: Johns Hopkins University Press.
Gilbreath, Dustin. 2015. “Are Georgian Politics Boring?” New Eastern Europe 1, no.
15: 105–11.
Giliomee, Hermann. 1995. “Democratization in South Africa.” Political Science Quarterly
110, no. 1: 83–104.
Ginsburg, Tom, and Aziz Z. Huq. 2018. How to Save a Constitutional Democracy.
Chicago: University of Chicago Press.
Gisselquist, Rachel M. 2008. “Democratic Transition and Democratic Survival in Benin.”
Democratization 15, no. 4: 789–814.
Gleditsch, Kristian Skrede. 2002. All International Politics Is Local: The Diffusion of
Conflict, Integration, and Democratization. Ann Arbor: University of Michigan Press.
Gleditsch, Kristian Skrede, and Michael D. Ward. 2006. “Diffusion and the International
Context of Democratization.” International Organization 60, no. 4 (October): 911–33.
280 References
Godio, Julio. 2002. Argentina, en la crisis está la solución: la crisis global desde las elecciones
de octubre de 2001 hasta la asunción de Duhalde. Buenos Aires: Editorial Biblos.
Godio, Julio. 2006. El tiempo de Kirchner. El devenir de una “revolución desde arriba.”
Buenos Aires: Ediciones Letra Grifa.
González-Ocantos, Ezequiel. 2016. Shifting Legal Visions: Judicial Change and Human
Rights Trials in Latin America. New York: Cambridge University Press.
González- Ocantos, Ezequiel A. 2020. The Politics of Transitional Justice in Latin
America: Power, Norms, and Capacity Building. Elements in politics and society in
Latin America. New York: Cambridge University Press.
Graham, Matthew H., and Milan W. Svolik. 2020. “Democracy in America? Partisanship,
Polarization, and the Robustness of Support for Democracy in the United States.”
American Political Science Review 114, no. 2 (May): 392–409. https://doi.org/10.1017/
S0003055420000052.
Guha, Ramchandra. 2007. India after Gandhi: The History of the World’s Largest Democracy.
New York: Harper Collins.
Guha, Ramachandra. 2017. India after Gandhi: The History of the World’s Largest Democracy.
10th ed. New Delhi: Picador.
Habib, Adam. 1995. “The Transition to Democracy in South Africa: Developing a Dynamic
Model.” Transformation 27: 50–73.
Habib, Adam, and Vishnu Padayachee. 2000. “Economic Policy and Power Relations
in South Africa’s Transition to Democracy.” World Development 28, no. 2
(February): 245–63.
Habib, Adam, Devan Pillay, and Ashwin Desai. 1998. “South Africa and the Global
Order: The Structural Conditioning of a Transition to Democracy.” Journal of
Contemporary African Studies 16, no. 1: 95–115. https://doi.org/10.1080/0258900980
8729622.
Haggard, Stephan, and Robert R. Kaufman. 2012. “Inequality and Regime
Change: Democratic Transitions and the Stability of Democratic Rule.” American
Political Science Review 106, no. 3 (August): 495–516.
Haggard, Stephan, and Robert R. Kaufman. 2016. Dictators and Democrats: Masses, Elites,
and Regime Change. Princeton, NJ: Princeton University Press.
Haggard, Stephen, and Robert Kaufmann. 2020. Democratic Backsliding. New York: Cambridge
University Press.
Hanson, Jonathan K., and Rachel Sigman. 2013. “Leviathan’s Latent Dimensions: Measuring
State Capacity for Comparative Political Research.” September 2013. Paper presented at
the APSA 2011 Annual Meeting. https://ssrn.com/abstract=1899933.
Harris, Vandra, and Andrew Goldsmith. 2011. “The Struggle for Independence Was Just
the Beginning.” In Security, Development and Nation-Building in Timor-Leste, edited by
Vandra Harris and Andrew Goldsmith, 3–16. London: Routledge.
Harris, Vandra, and Andrew O’Neil. 2011. “Timor-Leste’s Future(s): Security and Stability
2010–20.” In Security, Development and Nation-Building in Timor-Leste, edited by
Vandra Harris and Andrew Goldsmith, c hapter 13. London: Routledge.
Hashemi, Nader. 2009. Islam, Secularism, and Liberal Democracy: Toward a Democratic
Theory for Muslim Societies. New York: Oxford University Press.
Heilbrunn, John R. 1993. “Social Origins of National Conferences in Benin and Togo.”
Journal of Modern African Studies 31, no. 2: 277–99.
Helmke, Gretchen. 2005. Courts under Constraints: Judges, Generals, and Presidents in
Argentina. New York: Cambridge University Press.
References 281
Helsinki Watch. 1993. Human Rights in Moldova: The Turbulent Dniester. New York:
Human Rights Watch.
Herbst, Jeffrey. 2001. “Political Liberalization in Africa after Ten Years.” Comparative
Politics 33, no. 3: 357–75.
Herbst, Jeffrey. 2003. “Analyzing Apartheid: How Accurate Were US Intelligence
Estimates of South Africa, 1948–94?” African Affairs 102, no. 406: 81–107.
Heribert, Adam, and Kogila Moodley. 1993. The Opening of the Apartheid Mind: Options
for the New South Africa. Berkeley: University of California Press.
Herring, Ronald J. 1982. Land to the Tiller. New Haven, CT: Yale University Press.
Hicks, David. 2016. “Da coligação ao golpe: inevitabilidade e consequência na
descolonização do Timor Português.” In Timor-Leste: Colonialismo, Descolonização,
Lusotopia, edited by Rui Graça Feijó, 267–84. Lisboa: Edições Afrontamento.
Hofferbert, Richard I., and Hans-Dieter Klingemann. 1999. “Remembering the Bad
Old Days: Human Rights, Economic Conditions, and Democratic Performance in
Transitional Regimes.” European Journal of Political Research 36, no. 2: 155–74. https://
doi.org/10.1111/1475-6765.00466.
Hofstadter, Richard. 1969. The Idea of a Party System: The Rise of Legitimate Opposition in
the United States, 1780–1840. Berkeley: University of California Press.
Hood, Ludovic. 2006. “Security Sector Reform in East Timor, 1999–2004.” International
Peacekeeping 13, no. 1: 60–77.
Horowitz, Donald. 1985. Ethnic Groups in Conflict. Berkeley: University of California Press.
Horowitz, Donald. 2000. Ethnic Groups in Conflict. Updated edition with a new preface.
Berkeley: University of California Press.
Horowitz, Donald. 2013. Constitutional Change and Democracy in Indonesia.
Hough, Jerry. 1997. Democratization and Revolution in the USSR, 1985–1991. Washington,
DC: The Brookings Institution.
Houle, Christian. 2009. “Inequality and Democracy: Why Inequality Harms
Consolidation but Does Not Affect Democratization.” World Politics 61, no. 4
(October): 589–622.
Human Rights Watch. 2015. “Country Summary: Georgia.” Human Rights Watch,
January 2015. Accessed 2019. https://www.hrw.org/sites/default/files/related_mater
ial/georgia_5.pdf.
Human Rights Watch. 2019. World Report 2019. Accessed August 11, 2021. https://www.
hrw.org/world-report/2019#.
Huntington, Samuel P. 1968. Political Order in Changing Societies. New Haven, CT: Yale
University Press.
Huntington, Samuel P. 1991. The Third Wave: Democratization in the Late Twentieth
Century. Norman: University of Oklahoma Press.
IFES (International Foundation for Electoral Systems). 2018. “Election Guide: Democracy
Assistance and Election Guide.” Accessed 2018. https://www.electionguide.org.
ILO (International Labour Organization). 1975. ILO Yearbook 1975. Geneva: ILO.
ILO (International Labour Organization). 2002. Panorama laboral 2002: América Latina
y el Caribe. Lima: ILO.
Inglehart, Ronald. 1997. Modernization and Postmodernization: Cultural, Economic, and
Political Change in 43 Societies. Princeton, NJ: Princeton University Press.
Inglehart, Ronald, and Christian Welzel. 2005. Modernization, Cultural Change, and
Democracy: The Human Development Sequence. New York: Cambridge University Press.
282 References
Inglehart, Ronald, and Christian Welzel. 2007. “Mass Beliefs and Democratic Institutions.”
In Oxford Handbook of Comparative Politics, edited by Carles Boix and Susan Stokes,
297–316. New York: Oxford University Press.
Ingram, Sue. 2018. “Parties, Personalities and Political Power: Legacies of Liberal Peace-
Building in Timor-Leste.” Conflict, Security and Development 18, no. 5: 365–86.
Inman, Robert P., and Daniel L. Rubinfeld. 2012a. “Federal Institutions and the
Democratic Transition: Learning from South Africa.” Journal of Law, Economics, and
Organization 28, no. 4 (October): 783–817.
Inman, Robert P., and Daniel L. Rubinfeld. 2012b. “Understanding the Democratic
Transition in South Africa.” NBER Working Paper No. 17799, National Bureau of
Economic Research, February 2012. https://doi.org/10.3386/w17799.
International Policy Institute. 2003. A Review of Peace Operations: The Case for Change.
London: International Policy Institute, King’s College London.
Jaffrelot, Christophe, and Narender Kumar, eds. 2018. Dr. Ambedkar and Democracy: An
Anthology. Delhi: Oxford University Press.
Jaffrelot, Christophe, and Gilles Verniers. 2020, “A New Party System or a New Political
System?” Journal of Contemporary South Asia 28, no. 2: 141–54.
Jardine, Matthew. 1999. East Timor: Genocide in Paradise. Chicago: Odonian Press.
Jaunarena, Horacio. 2011. La casa está en orden: memoria de la transición. Buenos
Aires: Taeda Editora.
Jayal, Niraja Gopal. 2013. Citizenship and Its Discontents: An Indian History. Cambridge: Harvard
University Press.
Jensenius, Francesca. 2017. Social Justice through Inclusion. New York: Oxford University Press.
Jo, Hyeran. 2015. Compliant Rebels. New York: Cambridge University Press.
Jones, Mark P., and Wonjae Hwang. 2005. “Provincial Party Bosses: Keystone of
the Argentine Congress.” In Argentine Democracy: The Politics of Institutional
Weakness, edited by Steven Levitsky and Maria Victoria Murillo, 115–38. University
Park: Pennsylvania State University Press.
Jones, Mark P., and Juan Pablo Micozzi. 2013. “Argentina’s Unrepresentative and
Unaccountable Congress under the Kirchners.” In Representation and Effectiveness in
Latin American Democracies: Congress, Judiciary and Civil Society, edited by Moira B.
MacKinnon and Ludovico Feoli, 40–74. New York: Routledge.
Jones, Mark P., Sebastian Saiegh, Pablo T. Spiller, and Mariano Tommasi. 2002. “Amateur
Legislators–Professional Politicians: The Consequences of Party-Centered Electoral
Rules in a Federal System.” American Journal of Political Science 46, no. 3 (July): 656–69.
https://doi.org/10.2307/3088406.
Jowitt, Ken. 1992. New World Disorder: The Leninist Extinction. Berkeley: University of
California Press.
Kalashnyk, Pavlo. 2021. “A Blow to Political Opponents, a Message for Biden: What Is
Behind Zelenskyy’s Shutdown of Pro-Russian TV Channels?” Hromadske International,
February 16, 2021. https://en.hromadske.ua/posts/a-blow-to-political-opponents-a-
message-for-biden-what-is-behind-zelenskyys-shutdown-of-pro-russian-tv-channels.
Kalyvas, Stathis, and Laia Balcells. 2010. “International System and Technologies of
Rebellion: How the End of the Cold War Shaped Internal Conflict.” American Political
Science Review 104, no. 3 (August): 415–29.
Kammen, Douglas. 2009. “A Tape Recorder and a Wink? Transcript of the May 29, 1983,
Meeting between Governor Carrascalão and Xanana Gusmão.” Indonesia (Ithaca)
87: 73–102.
References 283
Kammen, Douglas. 2011. “The Armed Forces in Timor-Leste: Politicization through Elite
Conflict.” In The Political Resurgence of the Military in Southeast Asia, edited by Marcus
Mietzner, 107–26. London: Routledge.
Kammen, Douglas. 2015. Three Centuries of Conflict in East Timor. New Brunswick,
NJ: Rutgers University Press.
Kapstein, Ethan B., and Nathan Converse. 2008. The Fate of Young Democracies.
Karatnycky, Adrian, ed. 2002. Freedom in the World: The Annual Survey of Political Rights
and Civil Liberties, 2001–2002. New York: Hoepli Editore.
Karl, Terry Lynn. 1990. “Dilemmas of Democratization in Latin America.” Comparative
Politics 23, no. 1 (October): 1–21.
Karumidze, Zurab, and James V. Wertsch, eds. 2005. Enough! The Rose Revolution in the
Republic of Georgia 2003. New York: Nova Science Publishers.
Kedourie, Elie. 1994. Democracy and Arab Political Culture. Second edition with
corrections. London: Frank Cass.
Kenney, Charles D. 2004. Fujimori’s Coup and the Breakdown of Democracy in Latin
America. Notre Dame, IN: University of Notre Dame Press.
Khosla, Madhav, ed. 2014. Letters for a Nation: From Jawaharlal Nehru to His Chief
Ministers, 1947–1963. London: Allen Lane.
Khosla, Madhav. 2020. India’s Founding Moment: The Constitution of a Most Surprising
Democracy. Cambridge, MA: Harvard University Press.
King, Charles. 2000. The Moldovans: Romania, Russia, and the Politics of Culture. Palo
Alto, CA: Hoover Institution Press.
King, Dwight. 2003. “East Timor’s Founding Elections and Emerging Party System.”
Asian Survey 43, no. 5: 745–57.
Kingsbury, Damien. 2014. “Democratic Consolidation in Timor-Leste: Achievements,
Problems and Prospects.” Asian Journal of Political Science 22, no. 2: 181–205.
Kingsbury, Damien. 2018. “Timor- Leste’s Challenged Political Process: 2016– 17.”
Contemporary Southeast Asia 40, no. 1: 77–100.
Klug, Heinz. 2008. “South Africa’s Constitutional Court: Enabling Democracy and
Promoting Law in the Transition from Apartheid.” University of Wisconsin Law
School, Legal Studies Research Paper Series Paper No. 1530, 174–93. https://ssrn.com/
abstract=3521504.
Kohli, Atul, ed. 2001. The Success of India’s Democracy. New York: Cambridge University
Press.
Kohnert, Dirk, and Hans-Joachim Preuss. 2019. “Benin’s Stealthy Democracide: How
Africa’s Model Democracy Kills Itself Bit by Bit.” SSRN Electronic Journal. https://doi.
org/10.2139/ssrn.3481325.
Kopstein, Jeffrey S., and David A. Reilly. 2000. “Geographic Diffusion and the
Transformation of the Postcommunist World.” World Politics 53, no. 1: 1–37.
Kothari, Rajni, ed. 1970a. Caste in Indian Politics. Delhi: Orient Longman.
Kothari, Rajni. 1970b. Politics in India. Boston: Little, Brown and Company.
Kravchuk, Leonid. 2002. Maemo te, shcho maemo: Spohady i rozdumy. Kyiv, Ukraine:
Stolittia.
Krishnaswamy, Sudhir, and Siddharth Swaminathan. 2019. “Public Trust in the Indian
Judiciary: The Power to Transform.” In A Qualified Hope: The Indian Supreme Court
and Progressive Social Change, edited by Gerald N. Rosenberg, Sudhir Krishnaswamy,
and Shishir Bail, 123–47. Cambridge: Cambridge University Press.
284 References
Kudelia, Serhiy, and Taras Kuzio. 2014. “Nothing Personal: Explaining the Rise and
Decline of Political Machines in Ukraine.” Post-Soviet Affairs 31, no. 3: 250–78. http://
dx.doi.org/10.1080/1060586X.2014.920985.
Kuenzi, Michelle, and Gina Lambright. 2001. “Party System Institutionalization in 30
African Countries.” Party Politics 7, no. 4: 437–68.
Kushnarev, Evgeniy. 2007. Vybory i vily. Kyiv: ADEF-Ukraina.
Kvashilava, Bakur. 2019. “The Political Constraints for Civil Service Reform in
Georgia: History, Current Affairs, Prospects and Challenges.” Caucasus Survey 7, no.
3: 214–34.
LABORSTA Labour Statistics Database (Employment by economic activity). Accessed
July 2009. http://laborsta.ilo.org/.
Lachapelle, Jean, Steven Levitsky, and Lucan Way. 2020. “Social Revolution and
Authoritarian Durability.” World Politics 72, no. 4: 557–600.
Lang, Matt. 2009. Lineages of Despotism and Development: British Colonialism and State
Power. Chicago: University of Chicago Press.
Lanusse, Alejandro. 1988. “Entrevista al Tte. Gral. Alejandro Agustín Lanusse por
Gregorio A. Caro Figueroa.” Todo Es Historia 253: 58–63.
Larkins, Christopher. 1998. “The Judiciary and Delegative Democracy in Argentina.”
Comparative Politics 30, no. 4: 423–42. https://doi.org/10.2307/422332.
Leach, Michael. 2016. “FITUN: História preliminar de um movimento de resistência.” In
Timor-Leste: Colonialismo, descolonização, lusotopia, edited by Rui Graça Feijó, 397–
412. Lisboa: Edições Afrontamento.
Leach, Michael, and Damien Kingsbury. 2013. The Politics of Timor-Leste: Democratic
Consolidation after Intervention. Ithaca, NY: Cornell University Press.
Lebanidze, Bidzina. 2014. “What Makes Authoritarian Regimes Sweat? Linkage, Leverage
and Democratization in Post-Soviet South Caucasus.” Southeast European and Black
Sea Studies 14, no. 2: 199–218.
Lee, John. 2018. “Understanding Authoritarian Resilience and Countering Autocracy
Promotion in Asia.” Asia Policy 13, no. 4 (October): 99–122.
Lenz, Gabriel S. 2012. Follow the Leader? How Voters Respond to Politicians’ Policies and
Performance. Chicago: University of Chicago Press.
Lepsius, M. Rainer. 1978. “From Fragmented Party Democracy to Government
by Emergency Decree and Nazi Takeover.” In The Breakdown of Democratic
Regimes: Europe, edited by Juan J. Linz and Alfred Stepan, 34–79. Baltimore: Johns
Hopkins University Press.
Levitsky, Steven, and María Victoria Murillo. 2005. “Building Castles in the Sand? The
Politics of Institutional Weakness in Argentina.” In Argentine Democracy: The Politics
of Institutional Weakness, edited by Steven Levitsky and María Victoria Murillo, 21–44.
University Park: Pennsylvania State University Press.
Levitsky, Steven, and María Victoria Murillo. 2014. “Building Institutions on Weak
Foundations: Lessons from Latin America.” In Reflections on Uneven Democracies: The
Legacy of Guillermo O’Donnell, edited by Daniel Brinks, Marcelo Leiras, and Scott
Mainwaring, 189–213. Baltimore: Johns Hopkins University Press.
Levitsky, Steven, and Lucan A. Way. 2010. Competitive Authoritarianism: Hybrid Regimes
after the Cold War. New York: Cambridge University Press.
Levitsky, Steven, and Lucan A. Way. 2013. “The Durability of Revolutionary Regimes.”
Journal of Democracy 24, no. 3 (July): 5–17.
Levitsky, Steven, and Daniel Ziblatt. 2018. How Democracies Die. New York: Penguin Books.
References 285
Levitz, Philip, and Grigore Pop-Eleches. 2010. “Why No Backsliding? The EU’s Impact
on Democracy and Governance Before and After Accession.” Comparative Political
Studies 43, no. 4: 457–85.
Lieberman, Evan S. 2003. Race and Regionalism in the Politics of Taxation in Brazil and
South Africa. New York: Cambridge University Press.
Lijphart, Arend. 1977. Democracy in Plural Societies: A Comparative Exploration. New
Haven, CT: Yale University Press.
Lijphart, Arend. 1996. “The Puzzle of Indian Democracy: A Consociational
Interpretation.” American Political Science Review 90, no. 2: 258–68.
Lijphart, Arend. 1999. Patterns of Democracy: Government Forms and Performance in
Thirty-Six Countries. New Haven: Yale University Press.
Linz, Juan. 1978a. “The Breakdown of Democracy in Spain.” In The Breakdown of
Democratic Regimes: Europe, edited by Juan J. Linz and Alfred Stepan, 142–215.
Baltimore: John Hopkins University Press.
Linz, Juan. 1978b. The Breakdown of Democratic Regimes: Crisis, Breakdown, and
Reequilibration. Baltimore: Johns Hopkins University Press.
Linz, Juan. 1988. “Legitimacy of Democracy and the Socioeconomic System.” In
Comparing Pluralist Democracies: Strains on Legitimacy, edited by Mattéi Dogan, 65–
97. Boulder, CO: Westview Press.
Linz, Juan. 1990. “The Perils of Presidentialism.” Journal of Democracy 1, no. 1: 51–69.
Linz, Juan, and Alfred Stepan. 1989. “Political Crafting of Democratic Consolidation
or Destruction: European and South American Comparisons.” In Democracy in the
Americas: Stopping the Pendulum, edited by Robert A. Pastor, 41–61. New York: Holmes
& Meier.
Linz, Juan, and Alfred Stepan. 1996. Problems of Democratic Transition and
Consolidation: Southern Europe, South America, and Post- communist Europe.
Baltimore: Johns Hopkins University Press.
Lipset, Seymour Martin. 1959. “Some Social Requisites of Democracy: Economic
Development and Political Legitimacy.” American Political Science Review 53, no. 1
(March): 69–105.
Llanos, Mariana. 1998. “El presidente, el Congreso y la política de privatizaciones en la
Argentina (1989–1997).” Desarrollo Económico 38, no. 151: 743–70.
Lok Sabha Secretariat. 1949a. “Constituent Assembly Legislative Debates, Volume
VII: Tuesday, the 4th January 1949.” Delhi: Lok Sabha Secretariat. http://loksabhaph.
nic.in/Debates/cadebatefiles/C04011949.html.
Lok Sabha Secretariat. 1949b. “Constituent Assembly Legislative Debates, Volume
XI: Wednesday, the 23rd November 1949.” Delhi: Lok Sabha Secretariat. http://loksabh
aph.nic.in/Debates/cadebatefiles/C23111949.html.
Loney, Hannah. 2018. In Women’s Words: Violence and Everyday Life during the Indonesian
Occupation of East Timor, 1975–1999. Brighton: Sussex Academic Press.
Loxton, James. 2018. “Introduction: Authoritarian Successor Parties Worldwide.” In Life
after Dictatorship: Authoritarian Successor Parties Worldwide, edited by James Loxton
and Scott Mainwaring, 1–49. New York: Cambridge University Press.
Loxton, James, and Scott Mainwaring, eds. 2018. Life after Dictatorship: Authoritarian
Successor Parties Worldwide. New York: Cambridge University Press.
Lührmann, Anna, Nils Düpont, Masaaki Higashijima, Yaman Berker Kavasoglu, Kyle L.
Marquardt, Michael Bernhard, Holger Döring, Allen Hicken, Melis Laebens, Staffan
I. Lindberg, Juraj Medzihorsky, Anja Neundorf, Ora John Reuter, Saskia Ruth-Lovell,
286 References
Keith R. Weghorst, Nina Wiesehomeier, JosephWright, Nazifa Alizada, Paul Bederke,

Lisa Gastaldi, Sandra Grahn, Garry Hindle, Nina Ilchenko, Johannes von Römer,
Steven Wilson, Daniel Pemstein, Brigitte Seim. 2020. Varieties of Party Identity and
Organization (V-Party) Dataset V1. Varieties of Democracy (V-Dem) Project. Accessed
June 2021. https://doi.org/10.23696/vpartydsv1.
Lührmann, Anna, and Staffan Lindberg. 2019. “A Third Wave of Autocratization Is
Here: What Is New About It?” Democratization 26, no. 7: 1095–1113.
Lührmann, Anna, Marcus Tannenberg, and Staffan I. Lindberg. 2018. “Regimes of the
World (RoW): Opening New Avenues for The Comparative Study of Political Regimes.”
Politics and Governance 6, no. 1: 60–77.
Lundahl, Mats, and Fredrik Sjöholm. 2007. A Year of Turmoil: Timor-Leste 2006–2007.
Stockholm: Sida Country Economic Report.
Lupu, Noam, and Rachel Beatty Riedl. 2013. “Political Parties and Uncertainty in Developing
Democracies.” Comparative Political Studies 46, no. 11 (November): 1339–65.
Lynch, Gabrielle, and Gordon Crawford. 2011. “Democratization in Africa 1990–
2010: An Assessment.” Democratization 18, no. 2: 275–310.
Lyons, Terrence. 2004. “Post- conflict Elections and the Process of Demilitarizing
Politics: The Role of Electoral Administration.” Democratization 11, no. 3: 36–62.
MacQueen, Norrie. 2015a. “United Nations Integrated Mission in Timor- Leste
(UNMIT).” In Oxford Handbook of United Nations Peacekeeping Operations, edited
by Joachim Koops, Norrie MacQueen, Thierry Tardy, and Paul Williams, 755–66.
MacQueen, Norrie. 2015b. “United Nations Mission of Support in East Timor
(UNMISET)” In Oxford Handbook of United Nations Peacekeeping Operations, ed-
ited by Joachim Koops, Norrie MacQueen, Thierry Tardy, and Paul Williams, 683–93.
Madison, James. 1788. Federalist No. 51: “The Structure of the Government Must Furnish
the Proper Checks and Balances Between the Different Departments.” New York
Packet, February 8.
Madrid, Raúl L., and Kurt Weyland. 2019. “Conclusion: Why US Democracy Will Survive
Trump.” In When Democracy Trumps Populism: European and Latin American Lessons
for the United States, edited by Kurt Weyland and Raúl L. Madrid, 154–86.: Cambridge
University Press.
Magalhães, António Barbedo, Soei Liong Liem, and David Scott. 2007. Timor-
Leste: Interesses internacionais e actores locais. Porto: Edições Afrontamento.
Magaloni, Beatriz. 2006. Voting for Autocracy: Hegemonic Party Survival and Its Demise in
Mexico. New York: Cambridge University Press.
Magaloni, Beatriz. 2008. “Credible Power Sharing and the Longevity of Authoritarian
Rule.” Comparative Political Studies 41, nos. 4–5: 715–41.
Magnusson, Bruce A. 1999. “Testing Democracy in Benin: Experiments in Institutional
Reform.” In State, Conflict, and Democracy in Africa, edited by Richard Joseph, 217–37.
Boulder, CO: Lynne Rienner Publishers.
Magnusson, Bruce A. 2001. “Democratization and Domestic Insecurity: Navigating the
Transition in Benin.” Comparative Politics 33, no. 2: 211–30.
Magnusson, Bruce A. 2005. “Democratic Legitimacy in Benin: Institutions and Identity in a
Regional Context.” In The Fate of Africa’s Democratic Experiments: Elites and Institutions,
edited by Leonardo A. Villalón and Peter Von Doepp, 75–95. Bloomington: Indiana
University Press.
References 287
Magnusson, Bruce A., and John F. Clark. 2005. “Understanding Democratic Survival and
Democratic Failure in Africa: Insights from Divergent Democratic Experiments in Benin
and Congo (Brazzaville).” Comparative Studies in Society and History 47, no. 3: 552–82.
Mainwaring, Scott. 1989. “Transitions to Democracy and Democratic Consolidation: Theoretical
and Comparative Issues.” Kellogg Institute Working Paper #130, November 1989. https://kell
ogg.nd.edu/sites/default/files/old_files/documents/130_0.pdf.
Mainwaring, Scott, and Fernando Bizzarro. 2020. “Outcomes after Transitions in Third
Wave Democracies.” In The SAGE Handbook of Political Science, edited by Dirk Berg-
Schlosser, Bertrand Badie and Leonardo Morlino, 1540–57. London: SAGE Publications.
Mainwaring, Scott, and Aníbal Pérez-Liñán. 2013. Democracies and Dictatorships in Latin
America: Emergence, Survival, and Fall. New York: Cambridge University Press.
Mainwaring, Scott, and Matthew Shugart. 1997. “Juan Linz, Presidentialism and
Democracy: A Critical Appraisal.” Comparative Politics 29, no. 4: 449–71.
Makgetla, Itumeleng, and Ian Shapiro. 2016. “Business and the South African Transition.”
Unpublished manuscript, Yale University.
Mamdani, Mahmood. 2002. “Amnesty or Impunity? A Preliminary Critique of the Report
of the Truth and Reconciliation Commission of South Africa (TRC).” Diacritics 32,
nos. 3–4: 33–59.
Mamdani, Mahmood. 2018. Citizen and Subject: Contemporary Africa and the Legacy of
Late Colonialism. Princeton, NJ: Princeton University Press.
Mampilly, Zachariah. 2011. Rebel Rulers: Insurgent Governance and Civilian Life during
War. Ithaca, NY: Cornell University Press.
Mann, Michael. 1984. “The Autonomous Power of the State: Its Origins, Mechanisms, and
Results.” Archives Européenes de Sociologie 25: 185–213.
Mann, Michael. 2004. The Dark Side of Democracy: Explaining Ethnic Cleansing.
Marais, Nel, and Jo Davies. 2015. “The Role of the Business Elite in South Africa’s
Democratic Transition: Supporting an Inclusive Political and Economic
Transformation.” Inclusive Political Settlements Paper no. 8.
March, Andrew. 2011. Islam and Liberal Citizenship: The Search for an Overlapping
Consensus. New York: Oxford University Press.
March, Luke, and Graeme P. Herd. 2006. “Moldova between Europe and Russia: Inoculating
against the Colored Contagion?” Post-Soviet Affairs 22, no. 4: 349–79.
Marinov, Nikolay, and Hein Goemans. 2013. “Coups and Democracy.” British Journal of
Political Science 44, no. 4 (October): 799–825.
McFaul, Michael. 1999. “The Perils of a Protracted Transition.” Journal of Democracy 10,
no. 2: 4–18.
Marsh, Christopher. 2005. “Russian Orthodox Christians and Their Orientation toward
Church and State.” Journal of Church and State 47, no. 3 (Summer): 545–61.
Marshall, Monty, Ted Gurr, and Keith Jaggers. 2016. “Polity IV Project: Political Regime
Characteristics and Transitions, 1800– 2015.” Accessed March 2019. http://www.
systemicpeace.org/inscrdata.html.
Marshall, Monty, and Keith Jaggers. 2018. “Polity IV Project: Political Regime
Characteristics and Transitions, 1800–2017.” Accessed April 2019. http://www.system
icpeace.org/inscrdata.html.
Marshall, Monty, and Keith Jaggers. 2020. “Polity V Project: Political Regime
Characteristics and Transitions, 1800–2018.” Center for Systemic Peace. Accessed
October 10, 2021. http://www.systemicpeace.org/inscr/p5v2018.xls.
288 References
Martin, Ian, and Alexander Mayer- Rieckh. 2005. “The United Nations and East
Timor: From Self-Determination to State-Building.” International Peacekeeping 12, no.
1: 104–20.
Masi, Andrés Alberto. 2014. Los tiempos de Alfonsín: la construcción de un liderazgo
democrático. Buenos Aires: Capital Intelectual.
Maslin Nir, Sarah. 2019. “It Was a Robust Democracy. Then the New President Took
Power.” New York Times, July 4, 2019. https://www.nytimes.com/2019/07/04/world/afr
ica/benin-protests-talon-yayi.html.
McCoy, Jennifer, Tahmina Rahman, and Murat Somer. 2018. “Polarization and the Global
Crisis of Democracy: Common Patterns, Dynamics, and Pernicious Consequences for
Democratic Polities.” American Behavioral Scientist 62, no. 1 (January): 16–42.
McFaul, Michael. 1999. “The Perils of a Protracted Transition.” Journal of Democracy 10,
no. 2: 4–18.
McFaul, Michael. 2002. “The Fourth Wave of Democracy and Dictatorship: Noncooperative
Transitions in the Postcommunist World.” World Politics 54, no. 2 (January): 212–44.
McGuire, James W. 1992. “Union Political Tactics and Democratic Consolidation
in Alfonsín’s Argentina, 1983– 1989.” Latin American Research Review 27, no. 1
(January): 37–74.
McGuire, James W. 1995a. “Interim Government and Democratic Consolidation: Argentina
in Comparative Perspective.” In Between States: Interim Governments and Democratic
Transitions, edited by Yossi Shain and Juan Linz, 179– 210. New York: Cambridge
University Press.
McGuire, James W. 1995b. “Political Parties and Democracy in Argentina.” In Building
Democratic Institutions: Party Systems in Latin America, edited by Scott Mainwaring
and Timothy Scully, 200–46. Stanford, CA: Stanford University Press.
McGuire, James W. 1997. Peronism without Perón: Unions, Parties, and Democracy in
Argentina. Stanford, CA: Stanford University Press.
McMahon, Edward R. 2002. “The Impact of US Democracy and Governance Assistance
in Africa: Benin Case Study.” Center on Democratic Performance, Working Paper
Series no. 118. Department of Political Science, Binghamton University, New York.
McMillan, Alastair, 2010. “The Election Commission.” In The Oxford Companion to
Politics in India, edited by Niraja Gopal Jayal and Pratap Bhanu Mehta, 98–116.
Delhi: Oxford University Press.
McWilliams, Andrew. 2008. “Customary Governance in East Timor.” In Democratic
Governance in Timor-Leste: Reconciling the Local and the National, edited by David
Mearns and Steven Farram, 129–42. Darwin: Charles Darwin University Press.
Mehta, Pratap Bhanu. 2003. The Burden of Democracy. Delhi: Penguin.
Mehta, Pratap Bhanu. 2007. “The Rise of Judicial Sovereignty.” Journal of Democracy 18,
no. 2: 70–83.
Mehta, Pratap Bhanu. 2019. “The Story of Indian Democracy Written in Blood and
Betrayal.” The Indian Express, August 6, 2019. https://indianexpress.com/article/opin
ion/columns/jammu-kashmir-article-370-scrapped-special-status-amit-shah-naren
dra-modi-bjp-5880797/.
Mehta, Uday Singh. 1999. Liberalism and Empire: A Study in Nineteenth-Century British
Liberal Thought. Chicago: University of Chicago Press.
Mendes, Nuno Canas. 2005. A “multidimensionalidade” da construção identitária em
Timor-Leste: Nacionalismo, estado e identidade nacional. Lisboa: Instituto Superior de
Ciências Sociais e Políticas.
References 289
Mickey, Robert. 2015. Paths Out of Dixie: The Democratization of Authoritarian Enclaves
in America’s Deep South, 1944–1972. Princeton, NJ: Princeton University Press.
Mietzner, Marcus, ed. 2012. The Political Resurgence of the Military in Southeast
Asia: Conflict and Leadership. London: Routledge.
Mignone, Emilio Fermín. 1986. Iglesia y dictadura: el papel de la Iglesia a la luz de sus
relaciones con el régimen militar. Buenos Aires: Ediciones del Pensamiento Nacional.
Mill, John Stuart. 1861. Considerations on Representative Government. London: Parker,
Son, and Bourn.
Mill, John Stuart. (1864) 1975. “Representative Government.” In Three Essays, by John
Stuart Mill, 143–423. Oxford: Oxford University Press.
Millo, Yiftach, and Jon Barnett. 2004. “Educational Development in East Timor.”
International Journal of Educational Development 24, no. 6: 721–37.
Mitchell, Lincoln. 2004. “Georgia’s Rose Revolution.” Current History 103, no. 675: 342–53.
Mitchell, Lincoln. 2013. “What’s Next for Georgia? The End of the Rose Revolution.”
World Affairs 175, no. 5 (January/February): 75–82.
Mitchell, Lincoln. 2016. “Rushing Towards Amending the Constitution.” Lincoln Mitchell
(blog). October 23, 2016. http://lincolnmitchell.com/georgia-analysis/2016/10/19/
m08m1831no97ojv yutcdzqm6st5jb2?rq=Rushing%20towards%20amending%20
the%20constitution.
Modi, Narendra. 2019. “Why India and the World Need Gandhi.” New York Times,
October 2, 2019. https://www.nytimes.com/2019/10/02/opinion/modi-mahatma-gan
dhi.html.
Montesquieu, Charles de Secondat. (1748) 1900. The Spirit of the Laws. Translated by
Thomas Nugent. New York: D. Appleton and Company.
Moore Jr, Barrington. 1966. Social Origins of Dictatorship and Democracy. Boston: Beacon Press.
Morjé Howard, Lise. 2007. UN Peacekeeping in Civil Wars. New York: Cambridge University Press.
Morjé Howard, Lise. 2008. UN Peacekeeping in Civil Wars. New York: Cambridge
University Press.
Mueller, Sean. 2014. “The Parliamentary and Executive Elections in the Republic of
Georgia, 2012.” Electoral Studies 34: 342–47.
Mukherjee, Aditya. 2015. “Nehru’s Legacy: Inclusive Democracy and People’s
Empowerment.” Economic and Political Weekly 1, no. 16 (April 18): 38–45.
Müller, Jan-Werner. 2016a. “Protecting Popular Self-Government from the People? New
Normative Perspectives on Militant Democracy.” Annual Review of Political Science 19
(May): 249–65. https://doi.org/10.1146/annurev-polisci-043014-124054.
Müller, Jan- Werner. 2016b. What Is Populism? Philadelphia: University of
Pennsylvania Press.
Muller, Martin. 2011. “Public Opinion Toward the European Union in Georgia.” Post-
Soviet Affairs 27, no. 1: 64–92.
Munyeka, Wiza. 2014. “The Relationship Between Economic Growth and Inflation in the
South African Economy.” Mediterranean Journal of Social Sciences 5, no. 15: 119–29.
Mustapic, Ana María. 2005. “Inestabilidad sin colapso. La renuncia de los
presidentes: Argentina en el año 2001.” Desarrollo Económico 45, no. 178: 263–80.
Myrttinen, Henri. 2013. “Resistance, Symbolism and the Language of Stateness in Timor‐
Leste.” Oceania 83, no. 3: 208–20.
Nash, Andrew. 2002. “Mandela’s Democracy.” In Thabo Mbeki’s World: The Politics and
Ideology of the South African President, edited by Sean Jacobs and Richard Calland,
243–56. London: Zed Books.
290 References
Natalucci, Ana. 2011. “Entre la movilización y la institucionalización: Los dilemas de los

movimientos sociales (Argentina, 2001–2010).” Polis (Santiago) 10, no. 28: 193–219.
https://doi.org/10.4067/S0718-65682011000100012.
Nathan, Laurie. 2004. “Accounting for South Africa’s Successful Transition to Democracy.”
Crisis States Discussion Paper Series DP5, June, London School of Economics, London.
NDI (National Democratic Institute). 2013. “Statement of the NDI Election Observer
Delegation to Georgia’s 2013 Presidential Election.” National Democratic Institute,
October 28, 2013. https://cesko.ge/res/docs/StatementonGeorgias2013PresidentialEl
ections.pdf.
NDI (National Democratic Institute). 2016. “Statement of the NDI International
Observation Mission to Georgia’s October 8, 2016 Parliamentary Election.” National
Democratic Institute, October 8, 2016. https://www.ndi.org/sites/default/files/NDI_
GEO_Parliamentary%20Elections_2016_Statement-ENG.pdf.
Nehru, Jawaharlal. 1942. Glimpses of World History. New York: John Day.
Nehru, Jawaharlal. 1946. The Discovery of India. New York: John Day.
Neundorf, Anja, Johannes Gerschewski, and Roman-Gabriel Olar. 2019. “How Do
Inclusionary and Exclusionary Autocracies Affect Ordinary People?” Comparative
Political Studies 53, no. 12: 1890–1925. https://doi.org/10.1177/0010414019858958.
Nichols, Jim. 2003. “Coup in Georgia [Republic]: Recent Developments and Implications.”
Report RS21685. Washington, DC: Congressional Research Service.
Niner, Sara, ed. 2000. To Resist Is to Win! The Autobiography of Xanana Gusmão.
Victoria: Aurora.
Niner, Sara. 2015. “Women in the Post-conflict Moment in Timor-Leste.” In Security,
Development and Nation-Building in Timor-Leste, edited by Vandra Harris and Andrew
Goldsmith, 55–72. London: Routledge.
Nodia, Ghia. 2005. “Georgia 2004.” In Nations in Transit, edited by Adrian Karatnycky
and Alexander J. Motyl, 234–53. New York: Freedom House.
Norden, Deborah L. 1996. Military Rebellion in Argentina: Between Coups and
Consolidation. Lincoln: University of Nebraska Press.
Norris, Pippa, and Max Grömping. 2019. “Electoral Integrity Worldwide.” The Electoral
Integrity Project, University of Sydney, May 2019. https://static1.squarespace.com/
static/58533f31bebafbe99c85dc9b/t/5ce60bd6b208fcd93be49430/1558580197717/
Electoral+Integrity+Worldwide.pdf.
Norris, Pippa, and Ronald Inglehart. 2019. Cultural Backlash: Trump, Brexit, and
Authoritarian Populism. New York: Cambridge University Press.
Novaro, Marcos. 1994. Pilotos de tormentas: crisis de representación y personalización de la
política en Argentina, 1989–1993. Buenos Aires: Ediciones Letra Buena.
Novaro, Marcos. 2002. “La Alianza, de la gloria del llano a la debacle del gobierno.” In El
derrumbe político en el ocaso de la convertibilidad, edited by Marcos Novaro, 31–105.
Buenos Aires: Grupo Editorial Norma.
Novaro, Marcos. 2006. Historia de la Argentina contemporánea. De Perón a Kirchner.
Buenos Aires: Edhasa.
Novaro, Marcos. 2010. Historia de la Argentina. 1955–2010. Buenos Aires: Siglo XXI.
Novaro, Marcos, Alejandro Bonvecchi, and Nicolás Cherny. 2014. Los límites de la
voluntad: los gobiernos de Duhalde, Néstor y Cristina Kirchner. Buenos Aires: Ariel.
Novaro, Marcos, and Vicente Palermo. 2006. La dictadura militar. Buenos Aires: Paidós.
Nwajiaku, Kathryn. 1994. “The National Conferences in Benin and Togo Revisited.”
Journal of Modern African Studies 32, no. 3 (September): 429–47.
References 291
Nygaard-Christensen, Maj. 2016. “Timor-Leste in 2015: Petro-Politics or Sustainable

Growth.” In Southeast Asian Affairs 2016, edited by Malcolm Cook and Daljit Singh
347–60. Singapore: ISEAS—Yusof Ishak Institute.
Nzouankeu, Jacques Mariel. 1993. “The Role of the National Conference in the Transition
to Democracy in Africa: The Cases of Benin and Mali.” Issue: A Journal of Opinion 21,
nos. 1–2: 44–50.
O’Brien, Thomas. 2017. “Unbuilding from the Inside: Leadership and Democratization in
South Africa and South Korea.” Government and Opposition 52, no. 4 (October): 614–39.
O’Donnell, Guillermo. 1973. Modernization and Bureaucratic-Authoritarianism: Studies
in South American Politics. Berkeley: Institute for International Studies, University of
California.
O’Donnell, Guillermo. 1983. “La cosecha del miedo.” Nexos 6, no. 6: 51–60. https://www.
nexos.com.mx/?p=4147.
O’Donnell, Guillermo. 1986. “Introduction to the Latin American Cases.” In Transitions
from Authoritarian Rule: Prospects for Democracy. Vol. 2, Latin America, ed-
ited by Guillermo O’Donnell, Philippe Schmitter, and Laurence Whitehead, 3–18.
O’Donnell, Guillermo. 1993. “On the State, Democratization and Some Conceptual
Problems: A Latin American View with Glances at Some Postcommunist Countries.”
World Development 21, no. 8 (August): 1355–69.
O’Donnell, Guillermo. 1994. “Delegative Democracy.” Journal of Democracy 5, no. 1
(January): 55–69.
O’Donnell, Guillermo, and Philippe Schmitter. 1986. Transitions from Authoritarian Rule.
O’Flaherty, J. Daniel. 1992. “Holding Together South Africa.” Foreign Affairs 72, no. 4: 126–36.
ODIHR (Office for Democratic Institutions and Human Rights). 1996. Presidential Elections
in the Republic of Moldova. November 17 and December 1, 1996. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2001. Republic of
Moldova: Parliamentary Elections, 25 February 2001. Final Report. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2005. Moldova: Parliamentary
Elections, 6 March 2005, Election Observation Mission Report. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2006. Final Report on the
26 March 2006 Parliamentary Elections in Ukraine. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2007. Ukraine: Early
Parliamentary Elections, 30 September 2007. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2008. Georgia: Extraordinary
Presidential Election, 5 January 2008, Final Report. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2010. Ukraine: Presidential
Election, 17 January and 7 February 2010. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2012. Georgia: Parliamentary
Elections 1 October 2012, OSCEODIHR Election Observer Mission Final Report. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2016. Presidential
Election, 30 October and 13 November (second round) 2016. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2017. Georgia: Parliamentary
Elections, 8 and 30 October 2016. Warsaw: OSCE.
ODIHR (Office for Democratic Institutions and Human Rights). 2019.
Georgia: Presidential Election 28 October and 28 November 2018, OSCEODIHR Election
Observer Mission Final Report. Warsaw: OSCE.
292 References
Ollier, María Matilde. 2009. De la revolución a la democracia: Cambios privados, públicos

y políticos de la izquierda argentina. Buenos Aires: Siglo XXI/Universidad Nacional de
San Martín.
Ologou, Expedit. 2017. “Constitutional Reform in Benin: Why Did It Fail So Far?” Paper
presented at the conference Recent Developments in Constitution-Building in and
Beyond Africa, Central European University in cooperation with IDEA and Friedrich
Ebert Stiftung, Budapest, July 11, 2017.
Onuch, Olga. 2014. “‘It’s the Economy, Stupid,’ or Is It? The Role of Political Crises
in Mass Mobilization: The Case of Argentina in 2001.” In Argentina since the 2001
Crisis: Recovering the Past, Reclaiming the Future, edited by C. Levey, D. Ozarow, and C.
Wylde, 89–113. New York: Palgrave Macmillan.
Orovec, Phillip, and Edward Holland. 2019. “The Georgian Dream? Outcomes from the
Summer of Protest.” Demokratizatsiya 27, no. 2: 249–57.
Oszlak, Oscar. 1984. “Introducción.” In “Proceso,” crisis y transición democrática, edited by
Oscar Oszlak, vol. 2:29–42. Buenos Aires: Centro Editor de América Latina.
Palermo, Vicente, and Marcos Novaro. 1996. Política y poder en el gobierno de Menem.
Buenos Aires: Grupo Editorial Norma.
Palmer, Lisa, and Andrew McWilliam. 2018. “Ambivalent ‘Indigeneities’ in an
Independent Timor-Leste.” Asia Pacific Viewpoint 59, no. 3: 265–75.
Pepinsky, Tom. 2019. “Religion, Ethnicity, and Indonesia’s 2019 Presidential Election.”
New Mandala, May 28, 2019. https://www.newmandala.org/religion-ethnicity-and-
indonesias-2019-presidential-election/.
Pérez Liñán, Aníbal. 2007. Presidential Impeachment and the New Political Instability in
Latin America. New York: Cambridge University Press.
Pérez-Liñán, Aníbal, and Scott Mainwaring. 2013. “Regime Legacies and Levels of
Democracy: Evidence from Latin America.” Comparative Politics 45, no. 4 (July): 379–97.
Pérez-Liñán, Aníbal, Nicolás Schmidt, and Daniela Vairo. 2019. “Presidential Hegemony
and Democratic Backsliding in Latin America, 1925– 2016.” Democratization 26
(4): 606–25.
Peuch, Jean-Christophe. 2004. “Georgia: Critics Say Police Violence, Media Intimidation
on the Rise.” Eurasia Insight, February 20.
Pevehouse, Jon C. 2005. Democracy from Above? Regional Organizations and
Democratization. New York: Cambridge University Press.
Popescu, Nicu. 2012. “Moldova’s Fragile Pluralism.” Russian Politics and Law 50, no. 4
(July-August): 37–50.
Powles, Anna, 2015. “Local Ownership in the Security Sector Space: Lessons from Timor-
Leste and the Solomon Islands.” In United Nations Peacekeeping Challenge, edited by
Anna Powles and Negar Partow, 203–16. London: Ashgate.
Prakash, Gyan. 2018. Emergency Chronicles: Indira Gandhi and Democracy’s Turning
Point. London: Penguin Random House.
Preuss, Hans-Joachim. 2019. “Suicide of a Democracy.” International Politics and Society
Journal, March 5, 2019. https://www.ips-journal.eu/regions/africa/suicide-of-a-
democracy-3438/.
Pridham, Geoffrey. 1991. “The Politics of the European Community, Transnational
Networks and Democratic Transition in Southern Europe.” In Encouraging
Democracy: The International Context of Regime Transition in Southern Europe, edited
by Geoffrey Pridham, 212–45. Leicester, UK: Leicester University Press.
References 293
Prodromou, Elizabeth. 2004. “Christianity and Democracy: The Ambivalent Orthodox.”

Journal of Democracy 15, no. 2 (April): 62–75.
Przeworski, Adam. 1991. Democracy and the Market: Political and Economic Reforms in
Eastern Europe and Latin America. Cambridge: Cambridge University Press.
Przeworski, Adam. 2005. “Democracy as an Equilibrium.” Public Choice 123, no. 3/4
(June): 253–73.
Przeworski, Adam. 2019. Crises of Democracy. New York: Cambridge University Press.
Przeworski, Adam, Michael Alvarez, José Antonio Cheibub, and Fernando Limongi.
2000. Democracy and Development: Political Institutions and Material Well-Being in the
World, 1950–1990. New York: Cambridge University Press.
Przeworski, Adam, and Fernando Limongi. 1997. “Modernization: Theories and Facts.”
World Politics 49, no. 2 (January): 155–83.
Przeworski, Adam, Susan C. Stokes, and Bernard Manin, eds. 1999. Democracy,
Accountability, and Representation. Cambridge: Cambridge University Press.
Pucciarelli, Alfredo Raúl. 2006. “La República no tiene ejército. El poder gubernamental y
la movilización popular durante el levantamiento militar de Semana Santa.” In Los años
de Alfonsín, ¿el poder de la democracia o la democracia del poder?, edited by Alfredo
Raúl Pucciarelli, 115–51. Buenos Aires: Siglo XXI.
Puglisi, Rosaria. 2014. “A Regional Perspective on Post-Maidan Domestic Security.”
Presentation at the 10th Annual Danyliw Seminar on Ukraine, University of Ottawa.
Putnam, Robert H. 1993. Making Democracy Work: Civic Traditions in Modern Italy.
Princeton, NJ: Princeton University Press.
Rabushka, Alvin, and Kenneth A. Shepsle. 1972. Politics in Plural Societies: A Theory of
Democratic Instability. Columbus, OH: Charles E. Merrill.
Rajamani, Lavanya, and Arghya Sengupta. 2010. “The Supreme Court of India: Promise,
Reach, and Overreach.” In The Oxford Companion to Politics in India, edited by Niraja
Gopal Jayal and Pratap Bhanu Mehta, 80–97. Delhi: Oxford University Press.
Rajendran, G. 1974. The Ezhava Community and Kerala Politics. Trivandrum: Academy
of Political Science.
Ramos-Horta, José. 1987. Funu: The Unfinished Saga of East Timor. Trenton, NJ: The Red
Sea Press.
Ranis, Peter. 1992. Argentine Workers: Peronism and Contemporary Class Consciousness.
Pittsburgh, PA: University of Pittsburgh Press.
Rantete, Johannes, and Hermann Giliomee. 1992. “Transition to Democracy through
Transaction? Bilateral Negotiations between the ANC and NP in South Africa.” African
Affairs 91, no. 365 (October): 515–42.
Reddy, K. Vivek. 2016. “Minority Educational Institutions.” In Oxford Handbook of the
Indian Constitution, edited by Sujit Choudhry, Madhav Khosla, and Pratap Bhanu
Mehta, 921–42. Oxford: Oxford University Press.
Reenock, Christopher, Jeffrey Staton, and Marius Radean. 2013. “Legal Institutions and
Democratic Survival.” Journal of Politics 75, no. 2 (April): 491–505.
Remmer, Karen. 1996. “The Sustainability of Political Democracy: Lessons from South
America.” Comparative Political Studies 29, no. 6 (December): 611–34.
Reynolds, Andrew. 2002. The Architecture of Democracy: Constitutional Design, Conflict
Management, and Democracy. Oxford: Oxford University Press.
Riedl, Rachel Beatty. 2014. Authoritarian Origins of Democratic Party Systems in Africa.
294 References
Riedl, Rachel Beatty. 2016. “Strong Parties, Weak Parties: Divergent Pathways to
Democracy in Sub- Saharan Africa.” In Parties, Movements and Democracy in
the Developing World, edited by Nancy Bermeo and Deborah Yashar, 122– 56.
Riedl, Rachel Beatty. 2018. “Authoritarian Successor Parties in Sub-Saharan Africa.” In
Life after Dictatorship: Authoritarian Successor Parties Worldwide, edited by James
Loxton and Scott Mainwaring, 175–228. New York: Cambridge University Press.
Riedl, Rachel Beatty, Dan Slater, Joseph Wong, and Daniel Ziblatt. 2020. “Authoritarian-
Led Democratization.” Annual Review of Political Science 23 (May): 315–32.
Roberts, Kenneth M. 1998. Deepening Democracy? The Modern Left and Social Movements
in Chile and Peru. Stanford, CA: Stanford University Press.
Robinson, Geoffrey. 2010. “If You Leave Us Here, We Will Die”: How Genocide Was Stopped
in East Timor. Princeton, NJ: Princeton University Press.
Roeder, Philip G. 1994. “Varieties of Post-Soviet Authoritarian Regimes.” Post-Soviet
Affairs 10, no. 1: 61–101.
Roll, Kate. 2014. “Inventing the Veteran, Imagining the State: Post-conflict Reintegration
and State Consolidation in Timor-Leste.” PhD diss., Department of Politics and
International Relations, Oxford University.
Roll, Kate. 2018. “Reconsidering Reintegration: Veterans’ Benefits as State-Building.”
In The Promise of Prosperity: Visions of the Future in Timor-Leste, edited by Judith
Bovensiepen, 139–53. Canberra, Australia: ANU Press.
Romero, Luis Alberto. 2006. “La democracia y la sombra del proceso.” In Argentina
1976–2006. Entre la sombra de la dictadura y el futuro de la democracia, edited by Hugo
Quiroga and César Tcach, 15–30. Rosario: Homo Sapiens.
Ronen, Dov. 1975. Dahomey: Between Tradition and Modernity. Ithaca, NY: Cornell
University Press.
Roper, Steven D. 2001. “From Semi-presidentialism to Parliamentarism: Constitutional
Change in Post-Soviet Moldova.” Paper presented at the American Political Science
Association meeting, San Francisco.
Rose, Richard, and William Mishler. 1996. “Testing the Churchill Hypothesis: Popular
Support for Democracy and Its Alternatives.” Journal of Public Policy 16, no. 1: 29–58.
https://doi.org/10.1017/S0143814X00007856.
Ross, Michael. 2001. “Does Oil Hinder Democracy?” World Politics 53, no. 3: 325–61.
Rudolph, Lloyd, and Susanne Rudolph. 1967. The Modernity of Tradition.
Rueschemeyer, Dietrich, Evelyn Huber Stephens, and John D. Stephens. 1992. Capitalist
Development and Democracy. Cambridge: Polity Press.
Rustow, Dankwart A. 1970. “Transitions to Democracy: Toward a Dynamic Model.”
Comparative Politics 2, no. 3 (April): 337–63.
Ryan, Alan. 2012. The Making of Modern Liberalism. Princeton, NJ: Princeton
University Press.
Saunders, Christopher C. 2009. “UN Peacekeeping in Southern Africa: Namibia, Angola
and Mozambique.” In From Global Apartheid to Global Village: Africa and the United
Nations, edited by Adekeye Adebajo, 269–81. Scottsville, South Africa: University of
KwaZulu-Natal Press.
Scambary, James. 2011. “Anatomy of a Conflict: The 2006–7 Communal Violence in East
Timor.” In Security, Development and Nation-Building in Timor-Leste, edited by Vandra
Harris and Andrew Goldsmith, 59–79. London: Routledge.
References 295
Schedler, Andreas. 2013. The Politics of Uncertainty: Sustaining and Subverting Electoral
Authoritarianism. Oxford: Oxford University Press.
Schenoni, Luis L., and Scott Mainwaring. 2019. “US Hegemony and Regime Change in
Latin America.” Democratization 26, no. 2: 269–87. https://doi.org/10.1080/13510
347.2018.1516754.
Schuster, Federico L., Germán J. Pérez, Sebastián Pereyra, Melchor Armesto, Martín
Armelino, Analía García, Ana Natalucci, Melina Vázquez, and Patricia Zipcioglu.
2006. “Transformaciones de la protesta social en Argentina 1989–2003.” Documentos
de Trabajo, Facultad de Ciencias Sociales, Universidad de Buenos Aires, no. 48. https://
dialnet.unirioja.es/servlet/articulo?codigo=1983812.
SEDLAC (CEDLAS and The World Bank). 2018. Socio-economic Database for Latin
America and the Caribbean. Accessed April 2021. https://www.cedlas.econo.unlp.edu.
ar/wp/en/estadisticas/sedlac/.
Seekings, Jeremy. 2009. “South Africa since 1994: Who Holds Power after Apartheid?” In
Turning Points in African Democracy, edited by Abdul Raufu Mustapha and Lindsay
Whitfield, 134–52. Woodbridge: Boydell & Brewer.
Seely, Jennifer 2007. “The Presidential Election in Benin, March 2006.” Electoral Studies
26: 196–200.
Seely, Jennifer. 2009. The Legacies of Transition Governments in Africa: The Cases of Benin
and Togo. Basingstoke: Palgrave Macmillan.
Seligson, Mitchell A. 2002. “The Renaissance of Political Culture or the Renaissance of the
Ecological Fallacy?” Comparative Politics 34, no. 3 (April): 273–92.
Senén González, Santiago, and Fabián Bosoer. 2009. Breve historia del sindicalismo
argentino. Buenos Aires: El Ateneo.
Seo, Sang-Hyun. 2008. “A Study on Democratic Transition in South Africa: Democratic
through Compromise and Institutional Choice.” PhD diss., University of South Africa.
Shani, Ornit. 2018. How India Became Democratic: Citizenship and the Making of Universal
Franchise. New York: Cambridge University Press.
Shaw, Mark. 1997. “State Responses to Organized Crime in South Africa.” Transnational
Organized Crime 3, no. 2 (Summer): 1–19.
Shiosaki, Elfie. 2017. “‘We Have Resisted, Now We Must Build’: Regionalism and Nation-
Building in Timor-Leste.” Journal of Southeast Asian Studies 48, no. 1: 53–70.
Shoesmith, Dennis. 2003. “Timor-Leste: Divided Leadership in a Semi-Presidential
System.” Asian Survey 43, no. 2: 231–52.
Shoesmith, Dennis. 2013. “Political Parties.” In The Politics of Timor-Leste, Democratic
Consolidation after Intervention, edited by Michael Leach and Damien Kingsbury,
121–43. Ithaca, NY: Cornell University Press.
Sidicaro, Ricardo. 2002. Los tres peronismos estado y poder económico. Buenos
Aires: Siglo XXI.
Sikkink, Kathryn. 2008. “From Pariah State to Global Protagonist: Argentina and the
Struggle for International Human Rights.” Latin American Politics and Society 50, no.
1: 1–29. https://doi.org/10.1111/j.1548-2456.2008.00002.x.
Sindre, Gyda Marås. 2016. “Internal Party Democracy in Former Rebel Parties.” Party
Politics 22, no. 4: 501–11.
Sing, Ming. 2010. “Explaining Democratic Survival Globally (1946–2002).” Journal of
Politics 72, no. 2 (April): 438–55.
Slater, Dan. 2015. “Democratic Disarticulation and Its Dangers: Cleavage Formation
and Promiscuous Powersharing in Indonesian Party Politics.” In Building Blocs: How
296 References
Parties Organize Society, edited by Cedric de Leon, Manali Desai, and Cihan Tugal,
123–50. Stanford: Stanford University Press.
Slater, Dan. 2018a. “After Democracy.” Foreign Affairs, November 6, 2018. https://www.for
eignaffairs.com/articles/2018-11-06/after-democracy.
Slater, Dan. 2018b. “Party Cartelization, Indonesian-Style: Presidential Power-Sharing
and the Contingency of Democratic Opposition.” Journal of East Asian Studies 18, no.
1 (March): 23–46.
Slater, Dan. 2020. “Violent Origins of Authoritarian Variation: Rebellion Type and Regime
Type in Cold War Southeast Asia.” Government and Opposition 55, no. 1: 21–40.
Slater, Dan, and Erica Simmons. 2013. “Coping by Colluding: Political Uncertainty and
Promiscuous Powersharing in Indonesia and Bolivia.” Comparative Political Studies 46,
no. 11 (November): 1366–93.
Slater, Dan, Benjamin Smith, and Gautam Nair. 2014. “Economic Origins of Democratic
Breakdown? The Redistributive Model and the Postcolonial State.” Perspectives on
Politics 12, no. 2 (June): 353–74.
Slater, Dan, and Maya Tudor. 2019. “Why Religious Tolerance Won in Indonesia but Lost
in India.” Foreign Affairs, July 3, 2019. https://www.foreignaffairs.com/articles/india/
2019-07-03/why-religious-tolerance-won-indonesia-lost-india.
Slater, Dan, and Joseph Wong. 2018. “Game for Democracy: Authoritarian Successor
Parties in Developmental Asia.” In Life after Democracy: Authoritarian Successor
Parties Worldwide, edited by James Loxton and Scott Mainwaring, 284– 313.
Smith, Anthony. 2004. “East Timor: Elections in the World’s Newest Nation.” Journal of
Democracy 15, no. 2 (April): 145–59.
Smith, Benjamin. 2007. Hard Times in the Lands of Plenty: Oil Politics in Iran and
Indonesia. Ithaca: Cornell University Press.
Snyder, Jack L. 2000. From Voting to Violence: Democratization and Nationalist Conflict.
New York: Norton.
Socor, Vladimir. 1992. “Moldavia Builds a New State.” RFE/RL Research Report 1, no. 1 (3
January): 42–45.
Socor, Vladimir. 2013. “Demise of Moldova’s Alliance for European Integration Surprises
European Union’s Leaders.” Eurasia Daily Monitor 10, no. 90.
Sousa, Ivo Carneiro de. 2001. “The Portuguese Colonization and the Problem of the East
Timorese Nationalism.” Lusotopie 2001: 183–94.
Southall, Roger. 2001. “Opposition in South Africa: Issues and Problems.” Democratization
8, no. 1 (September): 1–24.
Southall, Roger. 2018. “Polarization in South Africa: Toward Democratic Deepening or
Democratic Decay?” Annals of the American Academy of Political and Social Science
681, no. 1 (January): 194–208.
Spiller, Pablo T., and Mariano Tommasi. 2007. The Institutional Foundations of
Public Policy in Argentina: A Transactions Cost Approach. Cambridge: Cambridge
University Press.
Staniland, Paul. 2012. “Organizing Insurgency: Networks, Resources, and Rebellion in
South Asia.” International Security 37, no. 1 (Summer): 142–77.
Steinberg, David A. 2015. Demanding Devaluation: Exchange Rate Politics in the
Developing World. Ithaca and London: Cornell University Press.
Stepan, Alfred. 1988. Rethinking Military Politics: Brazil and the Southern Cone. Princeton,
NJ: Princeton University Press.
References 297
Stepan, Alfred, Juan Linz, and Yogendra Yadav. 2011. Crafting State-Nations: India and
Other Multinational Democracies. Baltimore: Johns Hopkins University Press.
Stepan, Alfred C., and Graeme B. Robertson. 2003. “An ‘Arab’ More than ‘Muslim’
Democracy Gap.” Journal of Democracy 14, no. 3 (July): 30–44.
Stepan, Alfred, and Cindy Skach. 1993. “Constitutional Frameworks and Democratic
Consolidation: Parliamentarianism versus Presidentialism.” World Politics 46, no. 1
(October): 1–22.
Steytler, Nico. 2005. “Local Government in South Africa: Entrenching Decentralised
Government.” In The Place and Role of Local Government in Federal Systems, edited by
Nico Steytler, 183–211. Johannesburg: Konrad-Adenauer-Stiftung.
Strating, Rebecca. 2016. Social Democracy in East Timor. Oxon: Routledge.
Stroh, Alexander. 2018. “Sustaining and Jeopardising a Credible Arbiter: Judicial
Networks in Benin’s Consolidating Democracy.” International Political Science Review
39, no. 5: 600–15.
Svampa, Maristella, and Sebastián Pereyra. 2003. Entre la ruta y el barrio: la experiencia de
las organizaciones piqueteras. Buenos Aires: Editorial Biblos.
Svolik, Milan. 2012. The Politics of Authoritarian Rule. New York: Cambridge University Press.
Svolik, Milan. 2015. “Which Democracies Will Last? Coups, Incumbent Takeovers, and
the Dynamic of Democratic Consolidation.” British Journal of Political Science 45, no.
4 (October): 715–38.
Svolik, Milan. 2019. “Polarization versus Democracy.” Journal of Democracy 30, no. 3
(July): 20–32.
Swenson, Geoffrey. 2018. “Legal Pluralism in Theory and Practice.” International Studies
Review 20, no. 3 (January): 438–62.
Tansey, Oisín. 2009. Regime-Building: Democratization and International Administration.
Tarrow, Sidney. 1996. “States and Opportunities: The Political Structuring of Social
Movements.” In Comparative Perspectives on Social Movements, edited by Doug McAdam,
John McCarthy, and Mayer Zald, 41–61. New York: Cambridge University Press.
Taylor, Brian. 2011. State Building in Putin’s Russia: Policing and Coercion after
Communism. New York: Cambridge University Press.
Taylor, John. 1999. East Timor: The Price of Freedom. London: Zed Books.
Tcach, César, and Hugo Quiroga, eds. 2006. Argentina, 1976–2006: entre la sombra de la
dictadura y el futuro de la democracia. Rosario: Homo Sapiens.
Teorell, Jan. 2010. Determinants of Democratization: Explaining Regime Change in the
World, 1972–2006. New York: Cambridge University Press.
The Economist Intelligence Unit. 2020. Democracy Index 2019: A Year of Democratic
Setbacks and Popular Protest. London: The Economist Intelligence Unit Limited.
Thiel, Hermann, and Robert B Mattes. 1998. “Consolidation and Public Opinion in South
Africa.” Journal of Democracy 9, no. 1: 95–110. https://doi.org/10.1353/jod.1998.0010.
Tocqueville, Alexis de. 1835. Democracy in America, vol. 1., trans. Henry Reeve.
London: Saunders and Otley.
Toft, Monica Duffy. 2012. “Self-Determination, Secession, and Civil War.” Terrorism and
Political Violence 24, no. 4: 581–600.
Torre, Juan Carlos. 1983. Los sindicatos en el gobierno, 1973–1976. Buenos Aires: Centro
Editor de América Latina.
Trochev, Alexei. 2010. “Meddling with Justice: Competitive Politics, Impunity and Distrusted
Courts in Post-Orange Ukraine.” Demokratizatsiya 18, no. 2 (Spring): 122–47.
298 References
Trochev, Alexei. 2011. “Editor’s Introduction.” Statutes and Decisions 46, no. 2 (March–
April): 4–7.
Tudor, Maya. 2013. The Promise of Power: The Origins of Democracy in India and Autocracy
in Pakistan. Cambridge: Cambridge University Press.
Tudor, Maya, and Dan Slater. 2016. “The Content of Democracy: Nationalist Parties and
Inclusive Ideologies in India and Indonesia.” In Parties, Movements and Democracy
in the Developing World, edited by Nancy Bermeo and Deborah J. Yashar, 28–60.
Tudor, Maya, and Dan Slater. 2020. “Nationalism, Authoritarianism, and Democracy: Historical
Lessons from South and Southeast Asia.” Perspectives on Politics 19, no. 3: 706–22. https://doi.
org/10.1017/S153759272000078X.
UCDP/PRIO Armed Conflict Dataset (v.18.1). Accessed August 2019. https://www.prio.
org/Data/Armed-Conflict/UCDP-PRIO/.
UNDP (United Nations Development Programme). 2002. Human Development Report
2002: Deepening Democracy in a Fragmented World. New York: Oxford University Press.
United Nations. 2006. Report of the Independent Special Committee of Inquiry for Timor-
Leste. Geneva, October 2. Accessed 2019. https://www.ohchr.org/Documents/Countr
ies/COITimorLeste.pdf.
US Department of State. 2002. 2001 Country Reports on Human Rights Practices: Argentina.
March 4. Accessed July 2021. https://2009-2017.state.gov/j/drl/rls/hrrpt/2001/wha/
8278.htm.
V-Dem Institute. 2021. Autocratization Turns Viral: Democracy Report 2021. Gothenburg: V-
Dem Institute.
Vachudova, Milada Anna. 2005. Europe Undivided: Democracy, Leverage and Integration
after Communism. Oxford: Oxford University Press.
Vachudova, Milada Anna. 2010. “Democratization in Postcommunist Europe: Illiberal
Regimes and the Leverage of the European Union.” In Democracy and Authoritarianism
in the Postcommunist World, edited by Valerie Bunce, Michael McFaul, and Kathryn
Stoner-Weiss, 82–104. New York: Cambridge University Press.
Valencia, Fabian, and Luc Laeven. 2012. “Systemic Banking Crises Database; An Update.”
IMF Working Papers 12/163, International Monetary Fund. https://ideas.repec.org/p/
imf/imfwpa/12-163.html.
Van de Walle, Nicolas. 2001. African Economies and the Politics of Permanent Crisis, 1979–
1999. New York: Cambridge University Press.
Van de Walle, Nicolas. 2003. “Presidentialism and Clientelism in Africa’s Emerging Party
Systems.” Journal of Modern African Studies 41, no. 2: 297–321.
van Peski, Caecilia. 2013. “Good Cop, Bad Cop: Georgia’s One Hundred Days of a New
Democratic Dream.” Security and Human Rights Monitor 24: 49–100.
Van Wyk, Barry. 2005. The Balance of Power and the Transition to Democracy in South
Africa. PhD diss., University of Pretoria.
Varshney, Ashutosh. 1995. Democracy, Development and the Countryside: Urban-Rural
Struggles in India. New York: Cambridge University Press.
Varshney, Ashutosh. 1998. “India Defies the Odds: Why Democracy Survives.” Journal of
Democracy 9, no. 3 (July): 36–50.
Varshney, Ashutosh. 2000. “Is India Becoming More Democratic?” Journal of Asian
Studies 59, no. 1 (February): 3–25.
Varshney, Ashutosh. 2002. Ethnic Conflict and Civic Life: Hindus and Muslims in India.
New Haven, CT: Yale University Press.
References 299
Varshney, Ashutosh. 2012. “Two Banks of the Same River? Social Orders and
Entrepreneurialism in India.” In Anxieties of Democracy: Tocquevillian Reflections on
India and the United States, edited by Partha Chatterjee and Ira Katznelson, 225–56.
Delhi: Oxford University Press.
Varshney, Ashutosh. 2013. Battles Half Won: India’s Improbable Democracy. Gurgaon,
Haryana, India: Penguin Books.
Varshney, Ashutosh. 2015. “Asian Democracy through an Indian Prism.” Journal of Asian
Studies 74, no. 4: 917–26.
Varshney, Ashutosh. 2019a. “Democratic Subversion.” The Indian Express, December 16.
Varshney, Ashutosh. 2019b. “The Emergence of Right-Wing Populism in India.” In Re-
Forming India: The Nation Today, edited by Niraja Gopal Jayal, 327–45. Delhi: Penguin.
Varshney, Ashutosh. 2019c. “Modi Consolidates Power: Electoral Vibrancy, Mounting
Liberal Deficits.” Journal of Democracy 30, no. 4: 63–77.
Verbitsky, Horacio. 1995. El vuelo. Buenos Aires: Planeta.
Vom Hau, Matthias. 2008. “State Infrastructural Power and Nationalism: Comparative
Lessons from Mexico and Argentina.” Studies in Comparative International
Development 43, nos. 3–4 (December): 334–54.
Walker, Ignacio. 1990. Socialismo y democracia: Chile y Europa en perspectiva comparada.
Santiago: CIEPLAN/Hachette.
Waters, Trevor. 1996. “Moldova: Continuing Recipe for Instability.” Jane’s Intelligence
Review 8, no. 9 (September 18): 398–401.
Way, Lucan. 2002. “Pluralism by Default in Moldova.” Journal of Democracy 13, no. 4
(October): 127–41.
Way, Lucan. 2015. Pluralism by Default: Weak Autocrats and the Rise of Competitive
Politics. Baltimore: Johns Hopkins University Press.
Way, Lucan. 2016. “The Authoritarian Threat: Weaknesses of Autocracy Promotion.”
Journal of Democracy 27, no. 1 (January): 64–75.
Way, Lucan. 2019. “Ukraine’s Post-Maidan Struggles: Free Speech in a Time of War.”
Journal of Democracy 30, no. 3 (July): 48–60.
Weiner, Myron. 1967. Party Building in a New Nation: The Indian National Congress.
Weiner, Myron. 1989. The Indian Paradox: Essays in Indian Politics. Newbury Park,
CA: Sage Publications.
Weiner, Myron. 2001. “The Struggle for Equality: Caste in Indian Politics.” In The
Success of India’s Democracy, edited by Atul Kohli, 193–225. New York: Cambridge
University Press.
Weldemichael, Awet Tewelde. 2013. Third World Colonialism and Strategies of
Liberation: Eritrea and East Timor Compared. New York: Cambridge University Press.
Welsh, David. 1993. “Domestic Politics and Ethnic Conflict.” Survival 35 (1): 63–80.
Welsh, David. 1994. “South Africa’s Democratic Transition.” The Brown Journal of World
Affairs 2, no. 1 (Winter): 221–30.
Welsh, David. 2009. The Rise and Fall of Apartheid. Charlottesville: University of
Virginia Press.
Welt, Cory. 2015. “The Curious Case of Rustavi-2: Protecting Media Freedom and the
Rule of Law in Georgia.” PONARS Policy Memo 400, November 15.
Welzel, Christian, and Ronald Inglehart. 2009. “Political Culture, Mass Beliefs, and Value
Change.” In Democratization, edited by Christian W. Haerpfer, Christian Welzel,
Patrick Bernhagen, and Ronald Inglehart, 126–44. Oxford: Oxford University Press.
300 References
Weyland, Kurt. 2013. “Latin America’s Authoritarian Drift: The Threat from the Populist
Left.” Journal of Democracy 24, no. 3 (July): 18–32.
Weyland, Kurt. 2020. “Populism’s Threat to Democracy: Comparative Lessons for the
United States.” Perspectives on Politics 18, no. 2 (June): 389–406.
Weyland, Kurt, and Raúl Madrid, eds. 2019. When Democracy Trumps Populism:
European and Latin American Lessons for the United States. New York: Cambridge
University Press.
Whitehead, Laurence. 1986. “International Aspects of Democratization.” In Transitions
from Authoritarian Rule: Comparative Perspectives, edited by Guillermo O’Donnell,
Philippe C. Schmitter, and Laurence Whitehead, 3–46. Baltimore: Johns Hopkins
University Press.
Wibbels, Erik, and Kenneth Roberts. 2010. “The Politics of Economic Crisis in Latin America.”
Studies in Comparative International Development 45, no. 4 (December): 383–409.
Wieczorek, Robert. 2012. “ANC Dominance and Democratic Consolidation in South
Africa.” Penn State Journal of International Affairs 1 (2): 27–38.
Wilkinson, Steven I. 2004. “India, Consociational Theory and Ethnic Violence.” In India
and the Politics of Developing Countries: Essays in Memory of Myron Weiner, edited by
Ashutosh Varshney, 150–77. Delhi: Sage Publications.
Wilson, Andrew. 2013. “Filat’s Gamble.” OpenDemocracy, May 23, 2013. https://www.
opendemocracy.net/en/odr/filats-gamble/.
Wilson, Bu V. E. 2012. “The Politics of Security-Sector Reform.” In The Politics of Timor-
Leste: Democratic Consolidation after Intervention, edited by Michael Leach, and
Damien Kingsbury, 179–96. Ithaca, NY: Cornell University Press.
Woodberry, Robert D. 2012. “The Missionary Roots of Liberal Democracy.” American
Political Science Review 106, no. 2 (May): 244–74.
World Development Indicators (GDP per capita growth (annual %) [NY.GDP.PCAP.
KD.ZG], accessed June 2020 and June 2021; GDP growth (annual %) [NY.GDP.MKTP.
KD.ZG], accessed January 2019 and July 2021; GDP per capita (constant 2010 US$)
[NY.GDP.PCAP.KD], accessed January 2019, June 2020, and June/July 2021; GDP
per capita (current US$) [NY.GDP.PCAP.CD], accessed June 2020; Inflation, GDP
deflator (annual %), accessed January 2019 and June 2021 [NY.GDP.DEFL.KD.ZG];
Total natural resources rents (% of GDP) [NY.GDP.TOTL.RT.ZS], accessed August
2015; GNI per capita, PPP (current international $) [NY.GNP.PCAP.PP.CD], accessed
August 2021; Gini index (World Bank estimate) [SI.POV.GINI], accessed August 2021;
Military expenditure (% of GDP) [MS.MIL.XPND.GD.ZS], accessed January 2019;
and Unemployment, total (% of total labor force) (national estimate) [SL.UEM.TOTL.
NE.ZS], accessed January 2019). https://databank.worldbank.org/source/world-deve
lopment-indicators.
Worldwide Governance Indicators (for control of corruption, government effectiveness,
and rule of law). Accessed June 2021. https://info.worldbank.org/governance/wgi/.
Wu, Tim. 2020. “What Really Saved the Republic from Trump?” New York Times,
December 10, 2020. https://www.nytimes.com/2020/12/10/opinion/trump-constitut
ion-norms.html.
Young, Crawford. 2012. The Postcolonial State in Africa: Fifty Years of Independence, 1960–
2010. Madison: University of Wisconsin Press.
Zakaria, Fareed. 1997. The Future of Freedom: Illiberal Democracy at Home and Abroad.
New York: W. W. Norton.
References 301
Zechmeister, Elizabeth J., and Noam Lupu. 2019. El pulso de la democracia. Nashville,
TN: LAPOP. https://www.vanderbilt.edu/lapop/ab2018/2018-19_AmericasBarometer_R
egional_Report_Spanish_W_03.27.20.pdf.
Ziblatt, Daniel. 2017. Conservative Parties and the Birth of Democracy. New York:
Cambridge University Press.
Zürcher, Christoph, Carrie L. Manning, Kristie D. Evenson, Rachel Hayman, Sarah Riese,
and Nora Roehner. 2013. Costly Democracy: Peacebuilding and Democratization after
War. Stanford: Stanford University Press.
Index
For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion,
appear on only one of those pages.
Note: Tables and figures are indicated by t and f following the page number
Aceh, 76, 85–86 economic crises in, 28–29, 188, 201–4

ADP (Moldova). See Agrarian Democratic ethnic fractionalization in, 19t
Party (Moldova) GDP growth in, 18–20, 19t, 191, 204,
AEI (Moldova). See Alliance for European 258, 259t
Integration (Moldova) GDP per capita in, 19t, 28, 191, 195, 199,
Africa, sub-Saharan 201–3, 204, 222
democratic forging and survival in, 95–98, Gini index in, 19t, 189–90, 198, 213, 223
97f, 122–23, 151–52 governing parties and coalitions in, 246f, 249
multipartism in, 96 government effectiveness in, 258–59, 259t
structural conditions in, 94, 95–96 human rights movements in, 197, 204,
African National Congress (South Africa). See 218–20, 242
ANC (South Africa) inflation in, 189–90, 195, 199, 201–3
Agrarian Democratic Party (Moldova), 136–37 institutional capacity in, 257t, 257–58
Alfonsín, Raúl judicial capacity in, 256–57, 256t
government of, 199, 201–3, 207, 211, 212, legislative powers in, 254, 255t
213, 214 the military in, 204–5, 212
and human rights abuses, 204–5 military rebellions in, 28–29, 188, 194, 200,
Alkatiri, Mari, 167, 171, 177 205, 212, 214, 215, 217, 242
Alliance for European Integration (Moldova), popular mobilizations in, 199–200, 210,
139–40, 145–46 215, 216
Amaral, Xavier do, 167, 168, 174 poverty and unemployment in, 28, 189–90,
Ambedkar, B. R., 55 199, 203–4
ANC (South Africa) state capacity of, 19t
and democratization, 107, 108–9, 110–13, See also Argentine democracy
114, 115, 124, 125, 228–29, 248 Argentine democracy
electoral performance of, 30, 106, 243 actors’ normative commitment to, 28, 30–31,
illiberalism score of, 233–36, 234f 184, 188–89, 192, 193, 194–95, 205–6,
internal reformulation of, 94–95, 115 223–24, 228–29, 233, 243–44
and the USSR, 110, 123 alternative explanations for, 220–23
Angola, 162f breakdowns of, 189–90, 199, 222
APODETI (Timor-Leste), 167–68, 170 breakdown probability of, 28, 29f
Arab League, 12t, 16, 17f and business associations, 28, 191, 200,
Argentina 216–18, 222, 242
1973–76 period in, 28, 188–89, 193, 199, 206, challenges to, 200–1
210, 215, 218, 220–22 citizens’ support for, 28, 215–16, 221, 223
1976–83 dictatorship in, 28–29, 188–89, and demise of extremist actors, 188–89,
195–96, 197, 198, 201–3, 204, 206–7 194–95, 218–19, 224
2001 crisis in, 28, 199–200, 203–4, 207, duration of, 18, 19t
208–10, 212, 215 and economic crisis, 28, 189, 190, 191–92, 194,
“dirty war” in, 204, 205, 212, 218 199–200, 201, 203, 205, 208–10, 218
304 Index
Argentine democracy (cont.) duration of, 18, 19t

and failure of 1976–83 dictatorship, 28, institutional framework of, 95, 120–21
188, 195, 217, 219, 228–29, 241–42, 243, and international actors, 116, 122–24
244, 262–63 liberal, 4f, 100–1, 104–5, 105f
and human rights memorialization, 188, and the National Conference, 107–8, 116–19,
197–98, 219, 228–29, 242 120–21, 122, 125, 152
and international influences, 219–20, 260 opposition elites in, 24, 94–95, 99–100, 107–8,
and the left, 188–89, 219–20 116, 117–19, 151–52
liberal, 4f, 201, 202f and pluralism by default, 152–54
and the military, 188–89, 200, 205, 212, 222 power-sharing in, 94–95, 99–101, 119–20,
and organized labor, 28, 191, 193, 200, 207, 122, 251
213–15, 242 redistribution in, 99, 100, 125
and the PJ, 28, 193, 206–10, 242 ruling elites’ bargaining in, 24, 94–95, 97,
survival of, 2–4, 28, 31–32, 189–90, 199, 98–100, 106–8, 116, 124, 125, 151–52
200–1, 223–24, 239t survival of, 3–4, 23–24, 94–95, 98, 99–100,
transition to, 188, 201 104–5, 108, 119, 120, 183–84, 239t
and the UCR, 28, 210–11 transition to, 97, 98, 104–5, 108, 115,
Armed Forces for the National Liberation of 116–17, 124
Timor-Leste. See FALINTIL (Timor-Leste) Bharatiya Janata Party. See BJP (India)
ASDT (Timor-Leste), 167, 174 BJP (India)
Asia Pacific electoral performance of, 62–63, 248
democracy in, 41, 42f, 43f illiberalism of, 21, 30, 229, 234f, 237,
electoral-liberal democracy gap in, 37, 39f 241, 263
Azerbaijan, 128, 132, 133 Bloc for a Democratic and Prosperous Moldova
(BDPM), 137–38
BDPM (Moldova). See Bloc for a Democratic Bolivia, 198, 260–61
and Prosperous Moldova (BDPM) Botswana, 41–46, 44f, 45f
Benin Brazil, democracy in, 1, 198, 219–20, 244
crisis in, 102, 104, 117 breakaway democracies, 164–66
ethnic fractionalization in, 18–20, 19t Burundi, 97f
ethno-regional divisions in, 94, 99, 101–2,
118–19, 188 Cape Verde, 162f, 252–53
GDP growth in, 19t, 258, 259t Carrascalão, João, 170, 175
GDP per capita in, 18–20, 19t, 23–24, Carrascalão, Mário Viegas, 167, 170, 175
102, 102f CGT (Argentina), 205–6, 213, 214, 215, 217.
Gini index of, 19t See also Argentine democracy: and
governing parties and coalitions in, 246f, 248 organized labor
government effectiveness in, 258–59, 259t Chile, 1, 219–20, 244
institutional capacity in, 257t, 257–58 China, 1–2
judicial capacity in, 256–57, 256t CNRT (Timor-Leste), 175
legislative powers in, 255t illiberalism score of, 234f, 237
military regime in, 101–2, 116 Cold War, 1–2, 84–85, 96, 130–31
state capacity in, 18–20, 19t, 152 Confederación General del Trabajo (Argentina).
See also Beninese democracy See CGT (Argentina)
Beninese democracy Congress Party (India)
actors’ normative commitment to, 229, 237 and democracy, 30–31, 40, 49, 53, 55, 228–29,
backsliding of, 94–95, 100–1, 105–6, 122, 233, 236–37, 243, 263
151, 153–54, 241, 252–53 electoral performance of, 53–54, 56, 61,
bandwagoning in, 94–95, 106, 119–20, 125 63, 248
breakdown probability of, 23f, 30 illiberalism score of, 234f, 236
challenges to, 100–1, 105–6, 252 and land reform, 52
and citizen mobilization, 104–5, 121 Costa Rica, 41–46, 44f, 45f, 221
and crisis, 98–99, 107–8, 115–16, 118–19 Côte d’Ivoire, 97f, 123–24
Index 305
Cowry Forces for an Emerging Benin. See FCBE and strong legislatures, 253–55, 264
(Benin) structural accounts of, 4, 7, 18, 32, 46–48
Czech Republic, 259–60 threats to, 6, 7–8, 74–75, 85, 93
and voters, 230, 261–62
De Klerk, F. W., 108–10 See also democratic breakdown, factors
De la Rúa, Fernando, 203, 208–10, 211, 221 associated with
democracy, illiberal, 74–75, 85, 92, 208, 231 democratization
democracy, liberal agent-centric explanations of, 5, 50–51
characteristics of, 192 and elite values and interests, 51
state of, 1–2, 6, 31–32, 73, 92 of European colonies, 51–52
democratic breakdown, factors associated with, pathways to, 96–98, 97f, 185
11–16, 12t, 15f, 73–74, 258 structural explanations of, 4, 18, 32
democratic consolidation, 73 Duhalde, Eduardo, 203, 207–8, 210, 211
and system of government, 58–59
democratic erosion, 31–32, 73–74, 263 East Asia
democratic moments democracy in, 11
and authoritarian weakness, 128–29, 132–34, Eastern Europe, 12t, 17f
150–51 democracy in, 11
defined, 128 East Timor. See Timor-Leste: as an Indonesian
emergence of, 132–35 province
and external pressure, 128–29, 134–35, 150–51 electoral authoritarianism, 22–23, 74–75, 85,
and old regime forces, 128–29, 134, 150–51 87–88, 90
Democratic Party (Timor-Leste). See PD European Union, 134–35, 230, 259–61
(Timor-Leste) Ezahavas. See India: caste system in
Democratic Party of Moldova (PDM), 139–40, 234f
democratic survival FALINTIL (Timor-Leste), 169, 174, 175
and actors’ normative preferences, 5, 6, 26–27, features of, 176–77
29–31, 32, 50, 74–75, 185, 190, 193–94, 228, Falkland/Malvinas war, 188, 195–96, 198, 204,
230, 232, 262 217, 218, 242
and actors’ self-interest, 5–6, 26–27, 190–92, FCBE (Benin), 153
193, 244–58 illiberalism score of, 234f
and class structure, 5, 46, 48 Fernández de Kirchner, Cristina, 198, 207–8,
and economic development, 5, 7–8 210, 222, 249
and economic inequality, 5, 8–9, 46, 47 FRETILIN (Timor-Leste)
and ethnic homogeneity, 9–10, 48–49 agenda of, 167–68
and external actors, 11, 230, 259–61 and democracy, 174, 175
and foreign aid, 180, 181t and the separatist war, 161, 168, 169, 170,
and GDP per capita, 8, 46–47, 258 173–74, 176–77
in hard places, 2–4, 6, 18–20, 31–32, 73, 186, FRTLI (Timor-Leste)
228, 244 illiberalism score of, 234f, 237
and historical inheritances, 73–74, 82, 85
and institutional constraints, 5, 29–30, 220–21, Gandhi, Indira
229–30, 244, 252, 263, 264 administration of, 60–61
and internal war, 164–66, 165f, 184, 185 suspension of democracy by, 21, 56–57, 58,
and international environment, 11 236, 237
and intertemporal bargains, 250–51 Gandhi, Mahatma, 53, 236, 243
and natural resources, 179 General Confederation of Labor (Argentina).
and partisan power balancing, 221, 245, 249, See CGT (Argentina)
250, 263 Georgia
and political culture, 5, 10–11, 46, 48, 261 elections in, 145, 146–48, 149
and religion, 10–11 ethnic fractionalization, 19t
and state capacity, 9 GDP growth, 19t, 258, 259t
and strong courts, 255–58, 264 GDP per capita, 19t, 24–26
306 Index
Georgia (cont.) Election Commission of, 36, 60–61, 63

Gini index, 19t elections in, 34–35, 56, 57, 62–63
governing parties and coalitions in, 246f, 248 elites in, 21, 52–53, 54
government effectiveness in, 258–59, 259t Emergency period in, 34–35, 36, 56–57,
institutional capacity in, 257t, 257–58 236, 237
judicial capacity in, 256–57, 256t ethnic fractionalization in, 18–20, 19t, 49,
legislative powers in, 255t 188, 252
the media in, 148–49 federalism in, 49–50
party weakness in, 133, 148, 149 GDP growth in, 19t, 258, 259t
pluralism by default in, 26 GDP per capita in, 18–20, 19t
state breakdown in, 131–32, 144 governing parties and coalitions in, 246f, 248
state capacity in, 19t government effectiveness in, 258–59, 259t
See also Georgian democracy Hindu nationalism in, 21, 36, 40, 62, 63–64,
Georgian democracy 67, 68
actors’ normative commitment to, 229 independence of, 34–35, 49, 52, 53, 54–55
breakdown probability of, 25f institutional capacity in, 257t, 257–58
democratic moments of, 24–26, 128, 130–32, judicial capacity in, 256t
135, 144, 147–49, 150 the judiciary in, 36, 57, 59–60, 66–67, 68
duration of, 18, 19t land reform in, 52–53
external pressures on, 135, 145, 146–47, 150–51 legislative powers in, 255t
illiberalism in, 229, 238, 239t lynching in, 65
liberal, 4f, 130–31 Muslims in, 40, 49, 63–64, 65, 66, 67
old regime forces in, 26, 134, 135, 148–49 populism in, 1, 21
the opposition in, 144, 145–47, 150 religion in, 40, 50, 63–64, 66, 252
Russian influence on, 26, 134–35, 145, state capacity in, 19t
154, 260 universal franchise in, 54–56, 64, 68
survival of, 3–4, 24–26 See also Indian democracy
Georgian Dream (Georgia), 145–46, Indian democracy
147, 149–50 actors’ normative commitment to, 30–31,
and democracy, 131–32, 147–49 229, 233, 243–44
illiberalism score of, 234f breakdown probability of, 20f, 20
Germany, 8, 10, 51, 163–64, 232, 254 electoral, 35–36, 37, 38f, 41–46, 42f, 44f, 68
Ghana, 96–97, 97f and elite values and interests, 52–56, 68, 184,
Gini index, 12t, 13, 15f 241, 243, 262–63
Golkar (Indonesia), 84–85, 87, 88 erosion of, 34–36, 37, 62, 63, 68, 151, 229,
Greece, 1, 259–60 241, 252, 253, 263
Guinea-Bissau, 162f and federalism, 61–62, 67, 251–52
Gusmão, Xanana, 171, 176–77 illiberalism in, 229
and democracy, 172–73, 174–75, 242–43 liberal, 4f, 10–11, 35–36, 37, 38f, 41–46, 43f,
government of, 178 45f, 49–50, 68, 164f
liberal deficits of, 35–36, 40, 49, 63, 64–65, 68
Habibie, Bacharuddin Jusuf, 79 liberal-electoral gap of, 37, 39f, 67, 68
Hungary, 1, 220–21, 241, 259–61 longevity of, 18, 19t, 34, 35, 36, 59, 68
patterns of, 40
INC (India). See Congress Party (India) and political parties, 61–62, 68
India survival of, 2–4, 21, 31–32, 184, 228–29, 239t
as a British colony, 52, 53, 54 Indian National Congress. See Congress Party
caste system in, 47–48 (India)
civil liberties in, 163f Indian People’s Party. See BJP (India)
class structure in, 48 Indonesia
Constituent Assembly of, 52–53, 55–56, 63–64 1998–99 crisis in, 77, 83–85, 86–87
Constitution of, 36, 52–53, 55, 56–58, 64–65, 68 civil liberties in, 163f
economic inequality in, 19t, 47 economic growth in, 19t, 77–79, 77f, 83–84
Index 307
elections in, 75, 84, 89–91 election of, 89–90

ethnic fractionalization in, 19t government of, 90–91, 92
ethnoreligious diversity in, 22, 76–77, 85–86, Justicialist Party (Argentina). See PJ (Argentina)
90–91, 188, 252
GDP growth in, 19t, 258, 259t Kashmir, 66–67
GDP per capita in, 19t, 22, 78f, 79 Kérékou, Mathieu
Gini index in, 19t and democracy, 100, 124
governing parties and coalitions in, 246f, 248 as elected president, 119, 121, 122, 152–53
government effectiveness in, 258–59, 259t as a figurehead, 117, 118
independence of, 80–81, 82–83 military regime of, 101–2, 117, 118–19,
institutional capacity in, 257t, 257–58 123–24, 151–52, 153
Islam in, 22, 76, 81, 92 Kirchner, Néstor, 198, 207–8, 210
Islamist terrorism in, 85–86 Kravchuk, Leonid, 138, 140
under Japanese rule, 81 Kuchma, Leonid, 133
judicial capacity in, 256t authoritarianism of, 134–35, 150–51, 154
legislative powers in, 255t government of, 140–41
the military in, 83, 86–87, 170
nationalism of, 80–81, 82, 91 Latin America
New Order regime in, 22–23, 75, 83–84 democracy in, 11, 41, 42f, 43f, 190, 223
population of, 22, 76 electoral-liberal democracy gap in, 37, 39f
property rights in, 170f Liberal Democratic Party of Moldova
state capacity in, 19t, 83–84 (PLDM), 139–40
See also Indonesian democracy; Timor-Leste: illiberalism score of, 234f
as an Indonesian province Lithuania, 1, 14
Indonesian Communist Party. See PKI (Indonesia) Lucinschi, Petru
Indonesian democracy election of, 133, 136–37
actors’ normative commitment to, 229 government of, 137–38
breakdown probability of, 22f
and egalitarian nationalism, 22–23, 74, 75, Madison, James, 229, 230
80–82, 85, 86, 90, 91, 172, 184, 251–52 Malaysia, 163, 163f, 164f
and electoral authoritarianism, 22–23, 75, Mali, 10–11, 14
87–90, 93 Mandela, Nelson
and elite collusion, 93n.2 and democracy in South Africa, 110–12, 114,
and illiberalism, 22–23, 75, 80, 85, 90, 91–93 124, 233–36
and institutional strength, 22–23, 74, 75, as a democrat, 107, 231
82–85, 88 Mauberism, 172
liberal, 4f, 79–80, 79f, 164f Mauritius, 41–46, 44f, 45f
longevity of, 18, 19t Megawati Sukarnoputri, 88, 89, 90
and the military, 22–23, 75, 86–87 Menem, Carlos, 197, 203, 207–8, 210, 212, 220,
and state capacity, 22–23, 75, 82–83, 86 233, 252
survival of, 2–4, 10–11, 22–23, 31–32, Middle East, democracy in, 11
184, 239t military takeover. See democratic survival:
transition to, 22, 74, 75, 76, 79, 84–86 threats to
Indonesian Democratic Party. See PDI Mill, John Stuart, 9–10, 48–49
(Indonesia) Modi, Narendra
Indonesian Democratic Party of Struggle. See election of, 34, 37, 60
PDIP (Indonesia) electoral performance of, 35, 62–63
Indonesian Nationalist Party. See PNI Hindu nationalism of, 36, 62, 63, 64, 67
(Indonesia) and the judiciary, 66–67
Ivanishvili, Bidzina, 145–47, 149 and lynchings, 65
partisanship of, 63
Jamaica, 41–46, 44f, 45f regime of, 35–36, 253
Jokowi (Joko Widodo) restrictive laws of, 65–66
308 Index
Moldova NP (South Africa). See National Party

competitive authoritarianism in, 131 (South Africa)
elections in, 136–37, 138–39 NU (Indonesia). See Nahdlatul Ulama
ethnic fractionalization in, 19t (Indonesia)
ethno-linguistic tensions in, 135–36
GDP growth in, 19t, 258, 259t OAS, 259–61
GDP per capita in, 19t Obama, Barack, 143–44, 146–47
Gini index, 19t OECD, liberal democracy in, 1–2, 2f
governing parties and coalitions in, 246f, 249 OIC, 12t, 14, 16, 17f
government effectiveness in, 258–59, 259t Organization for Economic Co-operation
institutional capacity in, 257–58, 257t and Development. See OECD, liberal
judicial capacity in, 256–57, 256t democracy in
legislative powers in, 254, 255t Organization of American States. See OAS
party weakness in, 133 Organization of Islamic Cooperation. See OIC
pluralism by default in, 26
repressive capacity of, 26, 133, 136–37 Pakistan, democratization of, 51–52, 71n.36
state capacity, 19t PAN (Indonesia), 88
See also Moldovan democracy Pancasila, 81, 86, 88, 89
Moldovan democracy Papua New Guinea, democracy in, 41–46, 44f,
actors’ normative commitment to, 229 45f, 163
breakdown probability of, 20, 26f, 30 parliamentary system. See democratic
democratic moments of, 24–26, 128, 130–31, consolidation: and system of government
135–36, 139–40, 150 Partai Golongan Karya (Indonesia). See Golkar
duration of, 18, 19t (Indonesia)
illiberalism in, 229, 238, 239t, 241 Parti de la Révolution Populaire du Bénin
liberal, 4f, 130–31 (PRPB). See Peoples’ Revolutionary Party
ruling party weakness in, 135–36 of Benin (PRPB)
Russian influence on, 26, 134–35, 154, 260 Partido Justicialista (Argentina). See PJ
survival of, 3–4, 24–26 (Argentina)
transition to, 131, 135–36 Party of Action and Solidarity (Moldova). See
Mozambique, 161, 162f PAS (Moldova)
Party of Communists of the Republic of
Nahdlatul Ulama (Indonesia), 88, 90–91 Moldova (PCRM), 138–39, 150–51
National Awakening Party (Indonesia). See PKB and democracy, 131, 134–35, 154
(Indonesia) illiberalism score of, 234f
National Congress for Timorese Reconstruction. Party of Functional Groups (Indonesia). See
See CNRT (Timor-Leste) Golkar (Indonesia)
nationalism Party of Regions (Ukraine), 142, 150–51
and democratization, 80, 81–82 illiberalism score of, 234f
National Mandate Party. See PAN (Indonesia) PAS (Moldova), 139–40
National Party (South Africa), 108–9, 113, 114, illiberalism score of, 234f
123, 243 PCRM (Moldova). See Party of Communists of
Nehru, Jawaharlal the Republic of Moldova (PCRM)
death of, 35, 56 PD (Timor-Leste), 175, 178
and democracy in India, 21, 53–55, 60, 231, PDI (Indonesia), 84, 87
236, 243 PDIP (Indonesia), 88, 89, 90, 91
New Order. See Indonesia: New Order in illiberalism score of, 234f, 238
normative preference for democracy PDM (Moldova). See Democratic Party of
defined, 192–93 Moldova (PDM)
origins and endurance of, 30–31, 241, 262–63 Peoples’ Revolutionary Party of Benin (PRPB),
North America 116–17, 118–19, 152
democracy in, 41, 42f, 43f Perón, Juan Domingo, 206, 207–8, 210–11, 213
electoral-liberal democracy gap in, 37, 39f Peronist Party (Argentina). See PJ (Argentina)
Index 309
Peru, 199, 249 Revolutionary Front for an Independent

Philippines, 76, 79, 163f, 164f Timor-Leste. See FRETILIN (Timor-Leste)
piqueteros, 210, 215–16 Russia, 1–3, 7–8, 9, 10–11, 14, 132, 133
PJ (Argentina)
and democracy, 206–10, 211, 217 Saakashvili, Mikheil, 131–32, 133, 144, 145,
illiberalism score of, 208, 209f, 233, 234f 146–47, 150–51
See also Argentine democracy: and the PJ São Tomé, 162f
PJ/FPV (Argentina). See PJ (Argentina) Senegal, 96–97, 97f
PKB (Indonesia), 88 Shevardnadze, Eduard, 144–45
illiberalism score of, 234f, 238 Slovakia, 1, 259–60
PKI (Indonesia), 78–79, 83, 84 Snegur, Mircea, 136–37
PLDM (Moldova). See Liberal Democratic Social Democrat Party. See PSD (Timor-Leste)
Party of Moldova (PLDM) Soglo, Nicéphore, 107, 119, 122, 152–53
pluralism by default, 26, 128–29, 132–33, 134, South Africa
151–52, 153 apartheid system in, 102–4, 107, 108–10,
PNI (Indonesia), 81 111–12
Poland, democracy in, 1, 219–21, 259–61 electoral system of, 113
political actors ethnic fractionalization in, 18–20, 19t
defined, 50, 51 federal system of, 99, 113–14
normative preferences of, 192–94, 230–32 GDP growth in, 18–20, 19t, 258, 259t
regime choices of, 5, 190–91 GDP per capita in, 19t, 102f, 103–4
Popular Democratic Association of Gini index in, 18–20, 19t, 23–24, 103–4
Timor (Timor-Leste). See APODETI governing parties and coalitions in, 246f, 248
(Timor-Leste) government effectiveness in, 258–59, 259t
Poroshenko, Petro, 143–44 institutional capacity in, 257t, 257–58
Portugal. See Timor-Leste: as a Portuguese judicial capacity in, 256t
colony legislative powers in, 255t
post-Soviet countries racial divisions in, 94, 99, 102–3
democratic stability in, 153–54 state capacity in, 19t
and democratization, 128, 129–30, 132, 150 See also South African democracy
party weakness in, 133 South African democracy
PPP (Indonesia), 84, 87 actors’ normative commitment to, 229,
Prabowo Subianto, 86–87, 89–90, 91 233, 243–44
presidential system. See democratic and anti-apartheid movement, 110–12,
consolidation: and system of 228–29, 233–36, 241, 243, 244, 262–63
government breakdown probability of, 24f
PRO (Argentina) challenges to, 100–1
illiberalism score of, 208, 209f, 233, 234f consensus model of, 97, 114, 125
PRPB (Benin). See Peoples’ Revolutionary Party and crisis, 98–99, 104, 108
of Benin duration of, 18, 19t
PSD (Timor-Leste), 175 institutional framework of, 95, 106, 125
and international actors, 122–24, 260
Radical Civic Union (Argentina). See UCR liberal, 4f, 100–1, 105f, 106
(Argentina) opposition elites in, 24, 94–95, 99–100, 112–13
Ramos-Horta, José, 172, 177–78 power-sharing in, 94–95, 99–100, 112–14,
RB (Benin). See Renaissance Party of 251–52
Benin (RB) redistribution in, 99, 100, 115, 125
Renaissance Party of Benin (RB), 119 ruling elites’ bargaining in, 24, 94–95, 97,
illiberalism score of, 234f 98–100, 106–7, 112–13, 115, 124, 125
Republican Proposal (Argentina). See PRO survival of, 2–4, 23–24, 31–32, 94–95, 98,
(Argentina) 99–100, 104–5, 108, 115, 183–84, 239t
Revolutionary Front for an Independent Timor. transition to, 97, 98, 107, 108, 110, 112–13,
See FRTLI (Timor-Leste) 151–52
310 Index
Soviet Union separatist war in, 2–3, 27, 30–31, 160–61,

collapse of, 96, 110, 123, 128, 133, 136, 138 166, 169–71, 173–74, 185, 242–43
Communist Party of the, 133, 136, 138, 140 state capacity of, 18–20, 19t
totalitarian system of, 129–30 See also Timorese democracy
See also post-Soviet countries Timorese democracy
Spain actors’ normative commitment to, 30–31,
democratic breakdown probability in, 20, 20f, 229, 233, 243–44
22f, 23f, 24f, 25f, 26f, 27f, 29f breakdown probability of, 27f, 27, 161
state failure. See democratic survival: threats to civil liberties in, 163–64
Suharto duration of, 18, 19t
anti-communist killings by, 22, 78, 83 and foreign aid, 180–83, 260
authoritarian regime of, 22, 75, 76, 78–79, 82, and inclusive mechanisms, 178
86–87, 88, 92 and internal war, 27–28, 30–31, 161, 166, 176,
institutional development by, 83–84 179, 180, 183, 184–85, 219, 228–29, 241,
Sukarno, 78, 81, 82–83 242–43, 244, 262–63
liabilities of, 160–61
Taiwan, 1 liberal, 4f, 163–64, 164f
Talon, Patrice, 94–95, 100–1, 105–6, 121–22, and natural resources, 179–80
123–24, 153, 238, 248, 252–53 Polity score of, 162f
Thailand, 76, 79, 163f, 163, 164f survival of, 2–4, 27–28, 31–32, 161–63, 179,
third wave of democracy, 3–4, 11, 77, 96 183–84, 239t
Timorese Democratic Union. See UDT Tocqueville, Alexis de, 10–11, 261
(Timor-Leste) Trump, Donald, 75, 231, 261
Timorese Social Democratic Association Tunisia, 10–11
(Timor-Leste). See ASDT (Timor-Leste) Turkey, 10–11, 92
Timor-Leste Turkmenistan, 128, 132, 133
1975 civil war in, 160, 166, 168, 176, 242–43 Tymoshenko, Yulia, 141, 142
anti-communism in, 167, 168
anti-democratic coalitions in, 28, 166, Ubaldini, Saúl, 213–15
169–71, 185 UCR (Argentina)
the Catholic Church in, 170–72 and democracy, 210–11, 217
elections in, 163–64, 173, 175, 177–78, 180 illiberalism score of, 208, 209f, 233, 234f
ethnic fractionalization in, 19t See also Argentine democracy: and the UCR
ethnolinguistic divisions in, 18–20, 160 UDT (Timor-Leste), 167–68, 170, 175
GDP growth in, 19t, 259t UIA (Argentina), 214, 217
GDP per capita in, 19t, 160, 161f, 179–80 Ukraine
Gini index in, 19t competitive authoritarianism in, 140–41
governing parties and coalitions in, 246f, 249 elections in, 141, 142, 143
government effectiveness in, 258–59, 259t ethnic fractionalization in, 19t
independence of, 86, 160, 166, 173 GDP growth in, 18–20, 19t, 258, 259t
as an Indonesian province, 27, 28, 76, 85–86, GDP per capita in, 19t
169–71 Gini index in, 19t
institutional capacity in, 257t, 257–58 governing parties and coalitions in,
judicial capacity in, 256–57, 256t 246f, 249
legislative powers in, 255t government effectiveness in, 258–59, 259t
the military in, 28, 166, 176–79, 185, 242–43 institutional capacity in, 257t, 257–58
national identity of, 28, 166, 167–68, 171–73, judicial capacity in, 256–57, 256t
176, 185, 242–43 legislative powers in, 255t
party system in, 28, 161, 166, 173–75, the media in, 142, 143–44
178, 185 old regime forces in, 26, 134, 135
as a Portuguese colony, 27–28, 166–68 party weakness in, 133
property rights in, 170f pluralism by default in, 26
security crisis in, 177–79, 182 repressive capacity in, 133–34
Index 311
state capacity in, 19t UNM (Georgia). See United National

See also Ukrainian democracy Movement (Georgia)
Ukrainian democracy Uruguay, 198, 219–20, 221, 244
backsliding of, 131, 229 USSR. See Soviet Union
breakdown probability of, 20, 25f, 30
democratic moments of, 24–26, 128, 130–31, Varieties of Democracy project, 1–2, 37
135, 141, 143–44, 150 measuring system of, 37, 72n.41
duration of, 18, 19t V-Dem. See Varieties of Democracy project
illiberalism of, 229, 238, 239t, 241 Venezuela, 203, 220–21, 249, 254–55, 260–61
liberal, 4f, 130–31
Orange Revolution in, 131, 140–41 Western Europe
Russian influence on, 26, 134–35, 154, 260 democracy in, 41, 42f, 43f
survival of, 3–4, 24–26, 183–84 electoral-liberal democracy gap in, 37, 39f
Unión Cívica Radical (Argentina). See UCR See also democratization, of European
(Argentina) colonies
Unión Industrial Argentina. See UIA (Argentina) West Papua, 76, 85–86
United Development Party (Indonesia). See
PPP (Indonesia) Yanukovych, Viktor
United Kingdom, 6. See also India: as a British authoritarianism of, 134–35
colony; Falkland/Malvinas war government of, 131, 141, 142
United National Movement (Georgia), 145, Yayi, Boni, 100, 105–6, 121, 122, 124,
148, 150 153, 253
United Nations, peace missions of, 180, 182 Yudhoyono, Susilo Bambang, 87, 89
United States of America Yushchenko, Viktor, 131, 133, 141–42
democracy in, 1
foreign policy of, 11, 198, 219, 223, 259–61 Zimbabwe, 97f

15 - Democracy in Hard Places-Oxford University Press (2022)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

15 - Democracy in Hard Places-Oxford University Press (2022)

Uploaded by

Copyright:

Available Formats

Democracy in Hard Places

Published in the United States of America by Oxford University Press

© Oxford University Press 2022

All rights reserved. No part of this publication may be reproduced, stored in

You must not circulate this work in any other form

Library of Congress Cataloging-in-Publication Data

List of Figures  vii

1. Introduction: Democracy in Hard Places  1

1.1. Democracy in the world, 1900–​present  2

3.2. Indonesian economic development levels, 1960–​2019  78

1.1. Regression Analysis of Correlates of Democratic Breakdown  12

Emilia Simison (MA Torcuato Di Tella University) is a PhD candidate in Political

Ashutosh Varshney (PhD Massachusetts Institute of Technology) is Sol Goldman

Lucan Ahmad Way (PhD University of California, Berkeley) is Professor of Political

ADP Agrarian Democratic Party

GNU Government of National Unity

PSD Social Democrat Party

If recent political events have taught us anything, it is that democracy is often

Liberal Democracy Index (V−Dem, v11)

OECD (population weighted)

Figure 1.1. Democracy in the world, 1900–​present

independence, checks on executive power, and electoral integrity—​have now

despite three protracted economic depressions, military mutinies, and a prior

case study details the constellation of obstacles to democracy faced by a given

1980 1990 2000 2010 2020

Argentina Benin Timor−Leste

Agents, Interests, and Regime Preferences

Agent-​centric accounts of political phenomena have sometimes taken the form

The remainder of this chapter proceeds as follows. First, we offer a review of

The Effects of Context

Democracy’s survival in “hard” places, and its looming retrenchment in com-

Table 1.1. Regression Analysis of Correlates of Democratic Breakdown

Random effects logit Penalized logit

(1) (2) (3) (4) (5) (6)

GDP (logged) −0.744*** −0.891* −1.372*** −0.720*** −0.589 −0.981***

Years since last −0.00623 0.137* 0.116 −0.00725 0.0783* 0.0537

is a dichotomous variable that captures a shift from electoral democracy to au-

Figure 1.3. Marginal effects of continuous predictors

experiences a probability of breakdown (0.28 percent) that is ten times as small

(b) Organization of Islamic Cooperation

(c) Eastern Europe

Figure 1.4. Marginal effects of qualitative predictors

Argentina 8,511.71 1.01 55.74 34 0.13 0.20 45.88 0 0 0

1978 1982 1987 1992 1997 2002 2007 2012 2017

Figure 1.5. India (1978–​2017), probability of democratic breakdown

1978 1982 1987 1992 1997 2002 2007 2012 2017

Figure 1.6. Indonesia (2000–​2017), probability of democratic breakdown

1978 1982 1987 1992 1997 2002 2007 2012 2017

Figure 1.7. Benin (1992–​2017), probability of democratic breakdown

1978 1982 1987 1992 1997 2002 2007 2012 2017

South Africa Spain

Figure 1.8. South Africa (1996–​2017), probability of democratic breakdown

1978 1982 1987 1992 1997 2002 2007 2012 2017

Figure 1.9. Georgia (2005–​2017), probability of democratic breakdown

1978 1982 1987 1992 1997 2002 2007 2012 2017

Figure 1.10. Ukraine (1995–​1997, 2007–​2013), probability of democratic

1978 1982 1987 1992 1997 2002 2007 2012 2017

Figure 1.11. Moldova (1995–​2004, 2010–​2017), probability of democratic

Figure 1.12. Timor-​Leste (2002–​2017), probability of democratic breakdown

1978 1982 1987 1992 1997 2002 2007 2012 2017

Figure 1.13. Argentina (1984–​2017), probability of democratic breakdown

to policy moderation: Argentina’s experience with a brutal and inept military

The Importance of Normative Commitments

List of Figures vii

1. Introduction: Democracy in Hard Places 1

1.1. Democracy in the world, 1900–present 2

3.2. Indonesian economic development levels, 1960–2019 78

1.1. Regression Analysis of Correlates of Democratic Breakdown 12

Figure 1.1. Democracy in the world, 1900–present

independence, checks on executive power, and electoral integrity—have now

Agent-centric accounts of political phenomena have sometimes taken the form

GDP (logged) −0.744*** −0.891* −1.372* −0.720* −0.589 −0.981***

Figure 1.5. India (1978–2017), probability of democratic breakdown

Figure 1.6. Indonesia (2000–2017), probability of democratic breakdown

Figure 1.7. Benin (1992–2017), probability of democratic breakdown

Figure 1.8. South Africa (1996–2017), probability of democratic breakdown

Figure 1.9. Georgia (2005–2017), probability of democratic breakdown

Figure 1.10. Ukraine (1995–1997, 2007–2013), probability of democratic

Figure 1.11. Moldova (1995–2004, 2010–2017), probability of democratic

Figure 1.12. Timor-Leste (2002–2017), probability of democratic breakdown

Figure 1.13. Argentina (1984–2017), probability of democratic breakdown

Figure 2.1. Electoral and liberal democracy index, India 1950–2020

Figure 2.2. The electoral-liberal democracy gap—decadal comparison

Let us now turn to some more cross-regional and cross-country comparisons.

If structural accounts are not fully adequate, where do we go next? Agent-centric

Stating further—and rightly—that “a country’s political regime is not simply

India’s post-independence elites began their political education in 1920, not

consecutive national elections—in 1952, 1957, and 1962—and the preeminent

Figure 3.1. Indonesian economic growth rates, 1961–2019

Figure 3.2. Indonesian economic development levels, 1960–2019

Figure 3.3. Indonesia’s liberal democracy score over time, 1953–2019

predation and postcolonial challenges in nation-building; and multilinguistic,

rights as primary objectives of reform. The anti-apartheid movement was

Two common perceptions of democratization and endurance in sub-Saharan